BitShares Forum
Other => Graveyard => BitShares PTS => Topic started by: dga on January 07, 2014, 11:59:36 am
-
Hi, all - I posted this on bitcointalk, but figured it's most relevant here. For those who don't follow things over there, I developed the improved Nvidia mining code for scrypt coins (LTC, etc.) last month.
I have a working prototype of an nvidia-based GPU miner for Protoshares, and I'm convinced that it's not going to utterly disrupt the ecosystem - the GPU is better than CPU, but not so much so that a good CPU won't compete. At least, until other people improve upon my code. Also, as you'll see below, mining with a GPU is very power efficient (and the $ mined per kilowatt-hour spent is better than for any other coin).
I don't want to go the route of long-term maintaining this or releasing closed-source builds that siphon off a portion of the mining income, though I like that model a lot. Instead, I'd like to ask for tips _now_ in advance of releasing it open source. I contacted the Protoshares company about their bounty, but heard nothing yet.
Is there a kickstarter-equivalent in bitcoin, or a recommended best way to do this?
Here are the numbers I can put together. Note that there's still work to be done to make this thing friendly. Right now it's about in the same state as the Nvidia miner code I released for scrypt/LTC. Gunky to compile, not platform friendly, etc.
Dual GTX 690s: 2400 c/s at 115W at 10% CPU.
650Ti: 330-340 c/s at 1% CPU.
Macbook Pro with GT 650M: 200 c/s.
Tesla K20c: ~800c/s.
As with my previous release, this may be Kepler-only, but I think the algorithm will work on Fermi. Part of why I want to open source it is to let the community build upon it. I know that there's more performance that could be tuned into it - probably on the order of 30% but maybe 200%. I also think it's likely that there's an existing GPU implemention running privately, and this release would level the field a bit.
Let me know what you all think.
-Dave
-
I think most people will need more profit before they do anything.
-
Hi, all - I posted this on bitcointalk, but figured it's most relevant here. For those who don't follow things over there, I developed the improved Nvidia mining code for scrypt coins (LTC, etc.) last month.
I have a working prototype of an nvidia-based GPU miner for Protoshares, and I'm convinced that it's not going to utterly disrupt the ecosystem - the GPU is better than CPU, but not so much so that a good CPU won't compete. At least, until other people improve upon my code. Also, as you'll see below, mining with a GPU is very power efficient (and the $ mined per kilowatt-hour spent is better than for any other coin).
I don't want to go the route of long-term maintaining this or releasing closed-source builds that siphon off a portion of the mining income, though I like that model a lot. Instead, I'd like to ask for tips _now_ in advance of releasing it open source. I contacted the Protoshares company about their bounty, but heard nothing yet.
Is there a kickstarter-equivalent in bitcoin, or a recommended best way to do this?
Here are the numbers I can put together. Note that there's still work to be done to make this thing friendly. Right now it's about in the same state as the Nvidia miner code I released for scrypt/LTC. Gunky to compile, not platform friendly, etc.
Dual GTX 690s: 2400 c/s at 115W at 10% CPU.
650Ti: 330-340 c/s at 1% CPU.
Macbook Pro with GT 650M: 200 c/s.
Tesla K20c: ~800c/s.
As with my previous release, this may be Kepler-only, but I think the algorithm will work on Fermi. Part of why I want to open source it is to let the community build upon it. I know that there's more performance that could be tuned into it - probably on the order of 30% but maybe 200%. I also think it's likely that there's an existing GPU implemention running privately, and this release would level the field a bit.
Let me know what you all think.
-Dave
Would you release this open source for 100 PTS?
-
Let me know what you all think.
-Dave
Would you release this open source for 100 PTS?
I think it's worth more (but then again, I'd say that. *grin*), but I want to do the right thing by the community, also (or I'd sell it privately as I suspect the previous people to do this did). How about half the earlier bounty paid for the non-implementation proof, which would be about 160 PTS at current exchange?
-Dave
-
Let me know what you all think.
-Dave
Would you release this open source for 100 PTS?
I think it's worth more (but then again, I'd say that. *grin*), but I want to do the right thing by the community, also (or I'd sell it privately as I suspect the previous people to do this did). How about half the earlier bounty paid for the non-implementation proof, which would be about 160 PTS at current exchange?
-Dave
If you produce it open source and at least one other person on this forum is able to compile and replicate your results then I will provide 160 PTS to have this in the open for everyone and maintain a fair playing field.
-
Let me know what you all think.
-Dave
Would you release this open source for 100 PTS?
I think it's worth more (but then again, I'd say that. *grin*), but I want to do the right thing by the community, also (or I'd sell it privately as I suspect the previous people to do this did). How about half the earlier bounty paid for the non-implementation proof, which would be about 160 PTS at current exchange?
-Dave
If you produce it open source and at least one other person on this forum is able to compile and replicate your results then I will provide 160 PTS to have this in the open for everyone and maintain a fair playing field.
OK.
What's the right approach for ensuring that we can both trust each other on this one? You have a verification mechanism by having someone replicate the results, which seems appropriate, but is there a similar surety on my side?
I'll start getting the source ready for a quick release now. I'll update it over the coming weekend to be easier to build and use.
-Dave
-
Let me know what you all think.
-Dave
Would you release this open source for 100 PTS?
I think it's worth more (but then again, I'd say that. *grin*), but I want to do the right thing by the community, also (or I'd sell it privately as I suspect the previous people to do this did). How about half the earlier bounty paid for the non-implementation proof, which would be about 160 PTS at current exchange?
-Dave
If you produce it open source and at least one other person on this forum is able to compile and replicate your results then I will provide 160 PTS to have this in the open for everyone and maintain a fair playing field.
OK.
What's the right approach for ensuring that we can both trust each other on this one? You have a verification mechanism by having someone replicate the results, which seems appropriate, but is there a similar surety on my side?
I'll start getting the source ready for a quick release now. I'll update it over the coming weekend to be easier to build and use.
-Dave
A first cut release is ready.
-
Let me know what you all think.
-Dave
Would you release this open source for 100 PTS?
I think it's worth more (but then again, I'd say that. *grin*), but I want to do the right thing by the community, also (or I'd sell it privately as I suspect the previous people to do this did). How about half the earlier bounty paid for the non-implementation proof, which would be about 160 PTS at current exchange?
-Dave
If you produce it open source and at least one other person on this forum is able to compile and replicate your results then I will provide 160 PTS to have this in the open for everyone and maintain a fair playing field.
OK.
What's the right approach for ensuring that we can both trust each other on this one? You have a verification mechanism by having someone replicate the results, which seems appropriate, but is there a similar surety on my side?
I'll start getting the source ready for a quick release now. I'll update it over the coming weekend to be easier to build and use.
-Dave
A first cut release is ready.
I've put all of the code except the .cu file on github to verify that this format for a release meets your expectations.
https://github.com/dave-andersen/cudapts
I've licensed my own parts of the code under Apache2. Because it's based on ptsminer, the .cpp and .hpp files are covered under whatever copyright they started out as, which seems to be lost in the Mac port that I started with. But the core important thing is the .cu code, and that's basically free for anyone to do anything with.
-Dave
-
I've put all of the code except the .cu file on github to verify that this format for a release meets your expectations.
https://github.com/dave-andersen/cudapts
I've licensed my own parts of the code under Apache2. Because it's based on ptsminer, the .cpp and .hpp files are covered under whatever copyright they started out as, which seems to be lost in the Mac port that I started with. But the core important thing is the .cu code, and that's basically free for anyone to do anything with.
-Dave
How about a contract signed by your PTS address, and I'll trust based on that and upload the .cu file?
-Dave
-
DGA,
Around here we don't do "contracts" but operate on the basis of trust and informal/non-binding agreements. My entire trust with the community and valuation of PTS, AGS, and future in the bitcoin space is on the line and I am very well known. So that is more of a guarantee than any signed contract, it is in my best interest to do right by everyone I deal with.
It is nice that it is based on ptsminer so does that mean it works with standard pools?
Post the code and when I hear feedback from those on the forum that it works I will pay you 160 PTS.
Dan
-
DGA,
Around here we don't do "contracts" but operate on the basis of trust and informal/non-binding agreements. My entire trust with the community and valuation of PTS, AGS, and future in the bitcoin space is on the line and I am very well known. So that is more of a guarantee than any signed contract, it is in my best interest to do right by everyone I deal with.
It is nice that it is based on ptsminer so does that mean it works with standard pools?
Post the code and when I hear feedback from those on the forum that it works I will pay you 160 PTS.
Dan
ponder.
Done.
It's on github now.
-
I have attempted to compile the code. However, I get the following errors under cygwin 64..
make -f makefile.mingw64
yasm -f win64 -D WINABI -o obj/sha512_avx.o_asm intel/sha512_avx.asm
yasm -f win64 -D WINABI -o obj/sha512_sse4.o_asm intel/sha512_sse4.asm
g++ -c -O2 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sha512.o_amd sha512.c
g++ -c -O1 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sph_sha2.o_a md sph_sha2.c
g++ -c -O1 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sph_sha2big. o_amd sph_sha2big.c
g++ -c -O2 -march=athlon64 -mmmx -msse -msse2 -mthreads -w -Wall -Wextra -Wform at -Wformat-security -Wno-unused-parameter -DWIN32 -D_WINDOWS -DBOOST_THREAD_US E_LIB -DBOOST_SPIRIT_THREADSAFE -I"/home/jng/cudapts-master/cudapts-master" -I" /usr/local/include" -o obj/main_poolminer.o_amd main_poolminer.cpp
main_poolminer.cpp:18:25: fatal error: sys/syscall.h: No such file or directory
#include <sys/syscall.h>
^
compilation terminated.
makefile.mingw64:66: recipe for target 'obj/main_poolminer.o_amd' failed
make: *** [obj/main_poolminer.o_amd] Error 1
i dont see a file in the github for syscall.h...also, how do I find the boost suffix?, I am almost there, so your help is appreciated.
-
i am getting the following error
balu@jagali:~/cudapts$ make -f makefile.unix
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
g++ -c -O2 -fpermissive -o obj/cpuid.o cpuid.c
yasm -f elf64 -o obj/sha512_avx.o intel/sha512_avx.asm
make: yasm: Command not found
make: *** [obj/sha512_avx.o] Error 127
-
DGA,
Around here we don't do "contracts" but operate on the basis of trust and informal/non-binding agreements. My entire trust with the community and valuation of PTS, AGS, and future in the bitcoin space is on the line and I am very well known. So that is more of a guarantee than any signed contract, it is in my best interest to do right by everyone I deal with.
It is nice that it is based on ptsminer so does that mean it works with standard pools?
Post the code and when I hear feedback from those on the forum that it works I will pay you 160 PTS.
Dan
ponder.
Done.
It's on github now.
Thanks, It looks like people are attempting to build it!
-
I have added a record to our transaction log for your payment, provide a PTS address and I will pay out once we have confirmation from the community that it works.
https://docs.google.com/a/invictus-innovations.com/spreadsheet/ccc?key=0AqTwk-e7yzJydFZ3bVVWT0o1OUwzXzdESHFBY0FkUWc&usp=drive_web#gid=0
-
I have attempted to compile the code. However, I get the following errors under cygwin 64..
make -f makefile.mingw64
yasm -f win64 -D WINABI -o obj/sha512_avx.o_asm intel/sha512_avx.asm
yasm -f win64 -D WINABI -o obj/sha512_sse4.o_asm intel/sha512_sse4.asm
g++ -c -O2 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sha512.o_amd sha512.c
g++ -c -O1 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sph_sha2.o_a md sph_sha2.c
g++ -c -O1 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sph_sha2big. o_amd sph_sha2big.c
g++ -c -O2 -march=athlon64 -mmmx -msse -msse2 -mthreads -w -Wall -Wextra -Wform at -Wformat-security -Wno-unused-parameter -DWIN32 -D_WINDOWS -DBOOST_THREAD_US E_LIB -DBOOST_SPIRIT_THREADSAFE -I"/home/jng/cudapts-master/cudapts-master" -I" /usr/local/include" -o obj/main_poolminer.o_amd main_poolminer.cpp
main_poolminer.cpp:18:25: fatal error: sys/syscall.h: No such file or directory
#include <sys/syscall.h>
^
compilation terminated.
makefile.mingw64:66: recipe for target 'obj/main_poolminer.o_amd' failed
make: *** [obj/main_poolminer.o_amd] Error 1
i dont see a file in the github for syscall.h...also, how do I find the boost suffix?, I am almost there, so your help is appreciated.
I am unable to build this as well I keep getting errors with CentOS and osfinder.sh..It seems to no be in the git golder so I downloaded another one but it still doesnt work. I get errors just trying to run the scrip by it self ..I also tried the no chrono version. THat seems to go much further but still fails.
Any help and I will make a contribution to your effort. Thanks
-
Hi in CentOs 6 you will need to:
Install Yasm: https://trac.ffmpeg.org/wiki/CentosCompilationGuide
The dependency of osfinder is here:
wget https://raw.github.com/thbaumbach/ptsminer/master/src/osfinder.sh
chmod +x osfinder.sh
And finally install Cuda
http://aspratyush.wordpress.com/2012/05/06/install-nvidia-cuda-on-centos-6/
I am installing now Cuda.. I will let you know hot it goes.
-
Simpler way to install Cuda
Download here:
http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-linux/index.html#package-manager-installation
Install rpm here
http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-linux/index.html#package-manager-installation
-
I managed to compile it but got this error when try to run it.
Initializing. Device has 965951488 free of 1341325312 total bytes of memory
Could not mmap hugepage, reverting to malloc: Cannot allocate memory
-
ryen what OS? and care to provide steps? thanks
-
i am getting the following error
balu@jagali:~/cudapts$ make -f makefile.unix
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
g++ -c -O2 -fpermissive -o obj/cpuid.o cpuid.c
yasm -f elf64 -o obj/sha512_avx.o intel/sha512_avx.asm
make: yasm: Command not found
make: *** [obj/sha512_avx.o] Error 127
You need to install yasm. On Ubuntu, apt-get install yasm. On CentOS, yum install yasm.
-Dave
-
I managed to compile it but got this error when try to run it.
Initializing. Device has 965951488 free of 1341325312 total bytes of memory
Could not mmap hugepage, reverting to malloc: Cannot allocate memory
You can ignore that error - it was probably working. :) I've committed an update that gets rid of it. It's just informational (and unimportant) with the way the code works now.
-Dave
-
I have added a record to our transaction log for your payment, provide a PTS address and I will pay out once we have confirmation from the community that it works.
https://docs.google.com/a/invictus-innovations.com/spreadsheet/ccc?key=0AqTwk-e7yzJydFZ3bVVWT0o1OUwzXzdESHFBY0FkUWc&usp=drive_web#gid=0
Thanks!
Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
-Dave
-
could this miner be emulated on this: http://code.google.com/p/gpuocelot/ ?
-
I managed to compile it but got this error when try to run it.
Initializing. Device has 965951488 free of 1341325312 total bytes of memory
Could not mmap hugepage, reverting to malloc: Cannot allocate memory
You can ignore that error - it was probably working. :) I've committed an update that gets rid of it. It's just informational (and unimportant) with the way the code works now.
-Dave
Hi Dave, with that error the miner wasn't working, not doing anything. I made a change in makefile.unix line 142
nvcc -O3 -c -arch=sm_35 -o $@ $<
changed to
nvcc -O3 -c -arch=sm_20 -o $@ $<
Recompiled and now running flawlessly no errors. I just tried my luck with that change to reflect the compute capability of my card. Mine is gtx 570. Getting about 670 c/m.
-
Hi Dave, is there anyway to control intensity (utilization) of the card like cgminer? Example I wanna watch youtube video while mine at same time. Now very jerky.
-
I managed to compile it but got this error when try to run it.
Initializing. Device has 965951488 free of 1341325312 total bytes of memory
Could not mmap hugepage, reverting to malloc: Cannot allocate memory
You can ignore that error - it was probably working. :) I've committed an update that gets rid of it. It's just informational (and unimportant) with the way the code works now.
-Dave
Hi Dave, with that error the miner wasn't working, not doing anything. I made a change in makefile.unix line 142
nvcc -O3 -c -arch=sm_35 -o $@ $<
changed to
nvcc -O3 -c -arch=sm_20 -o $@ $<
Recompiled and now running flawlessly no errors. I just tried my luck with that change to reflect the compute capability of my card. Mine is gtx 570. Getting about 670 c/m.
Ahhhhh! Thank you - I'm really sorry, I left that in for testing. Your change was right. I've updated the makefile to use sm_30 as a default and will add this to the readme for now.
Glad it's working! That's a great rate from an older card.
-
Hi Dave, is there anyway to control intensity (utilization) of the card like cgminer? Example I wanna watch youtube video while mine at same time. Now very jerky.
Not yet. I'm creating a wishlist - see TODO in the git repository.
Also, for everyone:
I've restructured the git repository, added a better README, and simplified the command line options. You'll find it happier to use now than last night.
./cudapts <paymentaddr>
is enough to get most people running. If you can build it - which is still the tricky part. :)
-Dave
-
make -f makefile.unix
g++ -c -O2 -fpermissive -o obj/cpuid.o cpuid.c
yasm -f elf64 -o obj/sha512_avx.o intel/sha512_avx.asm
yasm -f elf64 -o obj/sha512_sse4.o intel/sha512_sse4.asm
g++ -c -O2 -fpermissive -o obj/sha512.o sha512.c
g++ -c -O1 -fpermissive -o obj/sph_sha2.o sph_sha2.c
g++ -c -O1 -fpermissive -o obj/sph_sha2big.o sph_sha2big.c
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
g++ -c -O2 -pthread -DBOOST_SPIRIT_THREADSAFE -D_FILE_OFFSET_BITS=64 -I/root/cudapts/cudapts/src -I/root/cudapts/cudapts/src/obj -fno-stack-protector -fstack-protector-all -Wstack-protector -D_FORTIFY_SOURCE=2 -o obj/main_poolminer.o main_poolminer.cpp
In file included from main_poolminer.cpp:15:
main_poolminer.hpp:6:28: error: boost/thread.hpp: No such file or directory
main_poolminer.hpp:7:26: error: boost/asio.hpp: No such file or directory
main_poolminer.hpp:8:56: error: boost/date_time/posix_time/posix_time_io.hpp: No such file or directory
In file included from main_poolminer.cpp:15:
main_poolminer.hpp: In function ‘void print256(const char*, uint32_t*)’:
main_poolminer.hpp:54: error: aggregate ‘std::stringstream ss’ has incomplete type and cannot be defined
main_poolminer.hpp:56: error: ‘setw’ is not a member of ‘std’
main_poolminer.hpp:56: error: ‘setfill’ is not a member of ‘std’
main_poolminer.cpp: At global scope:
main_poolminer.cpp:40: error: ‘boost’ has not been declared
main_poolminer.cpp:40: error: expected initializer before ‘*’ token
main_poolminer.cpp:41: error: ‘boost’ has not been declared
main_poolminer.cpp:41: error: expected initializer before ‘t_start’
main_poolminer.cpp:126: error: ‘boost’ has not been declared
main_poolminer.cpp:126: error: ISO C++ forbids declaration of ‘shared_mutex’ with no type
main_poolminer.cpp:126: error: expected ‘;’ before ‘_mutex_getwork’
main_poolminer.cpp: In member function ‘virtual blockHeader_t* CBlockProviderGW::getBlock(unsigned int, unsigned int, unsigned int)’:
main_poolminer.cpp:69: error: ‘boost’ has not been declared
main_poolminer.cpp:69: error: ‘boost’ has not been declared
main_poolminer.cpp:69: error: ‘_mutex_getwork’ was not declared in this scope
main_poolminer.cpp:69: error: ‘lock’ was not declared in this scope
main_poolminer.cpp: In member function ‘virtual void CBlockProviderGW::setBlockTo(blockHeader_t*)’:
main_poolminer.cpp:89: error: ‘boost’ has not been declared
main_poolminer.cpp:89: error: ‘boost’ has not been declared
main_poolminer.cpp:89: error: ‘_mutex_getwork’ was not declared in this scope
main_poolminer.cpp:89: error: ‘lock’ was not declared in this scope
main_poolminer.cpp: In member function ‘virtual void CBlockProviderGW::submitBlock(blockHeader_t*, unsigned int)’:
main_poolminer.cpp:111: error: ‘socket_to_server’ was not declared in this scope
main_poolminer.cpp:115: error: ‘boost’ has not been declared
main_poolminer.cpp:115: error: expected `;' before ‘submit_error’
main_poolminer.cpp:116: error: ‘boost’ has not been declared
main_poolminer.cpp:116: error: ‘boost’ has not been declared
main_poolminer.cpp:116: error: ‘boost’ has not been declared
main_poolminer.cpp:116: error: ‘submit_error’ was not declared in this scope
main_poolminer.cpp:119: error: ‘submit_error’ was not declared in this scope
main_poolminer.cpp: At global scope:
main_poolminer.cpp:137: error: ‘boost’ has not been declared
main_poolminer.cpp:137: error: ISO C++ forbids declaration of ‘shared_mutex’ with no type
main_poolminer.cpp:137: error: ‘shared_mutex’ declared as a ‘virtual’ field
main_poolminer.cpp:137: error: expected ‘;’ before ‘&’ token
main_poolminer.cpp:229: error: ‘boost’ has not been declared
main_poolminer.cpp:229: error: ISO C++ forbids declaration of ‘shared_lock’ with no type
main_poolminer.cpp:229: error: expected ‘;’ before ‘<’ token
main_poolminer.cpp:235: error: ‘boost’ has not been declared
main_poolminer.cpp:235: error: ISO C++ forbids declaration of ‘thread’ with no type
main_poolminer.cpp:235: error: expected ‘;’ before ‘_thread’
main_poolminer.cpp: In constructor ‘CWorkerThread::CWorkerThread(CMasterThreadStub*, unsigned int, CBlockProviderGW*)’:
main_poolminer.cpp:144: error: class ‘CWorkerThread’ does not have any field named ‘_working_lock’
main_poolminer.cpp:144: error: class ‘CWorkerThread’ does not have any field named ‘_thread’
main_poolminer.cpp: In member function ‘void CWorkerThread::mineloop()’:
main_poolminer.cpp:165: error: ‘boost’ has not been declared
main_poolminer.cpp:165: error: ‘boost’ has not been declared
main_poolminer.cpp: In member function ‘void CWorkerThread::run()’:
main_poolminer.cpp:215: error: ‘boost’ has not been declared
main_poolminer.cpp:215: error: ‘boost’ has not been declared
main_poolminer.cpp: In member function ‘void CWorkerThread::work()’:
main_poolminer.cpp:224: error: ‘_working_lock’ was not declared in this scope
main_poolminer.cpp:224: error: expected type-specifier before ‘boost’
main_poolminer.cpp:224: error: expected `;' before ‘boost’
main_poolminer.cpp: At global scope:
main_poolminer.cpp:416: error: ‘boost’ has not been declared
main_poolminer.cpp:416: error: ISO C++ forbids declaration of ‘shared_mutex’ with no type
main_poolminer.cpp:416: error: expected ‘;’ before ‘&’ token
main_poolminer.cpp:420: error: expected `;' before ‘private’
main_poolminer.cpp:428: error: ‘boost’ has not been declared
main_poolminer.cpp:428: error: ISO C++ forbids declaration of ‘shared_mutex’ with no type
main_poolminer.cpp:428: error: expected ‘;’ before ‘_mutex_master’
main_poolminer.cpp:429: error: ‘boost’ has not been declared
main_poolminer.cpp:429: error: ISO C++ forbids declaration of ‘shared_mutex’ with no type
main_poolminer.cpp:429: error: expected ‘;’ before ‘_mutex_working’
main_poolminer.cpp: In member function ‘void CMasterThread::run()’:
main_poolminer.cpp:246: error: ‘boost’ has not been declared
main_poolminer.cpp:246: error: ‘boost’ has not been declared
main_poolminer.cpp:246: error: ‘_mutex_master’ was not declared in this scope
main_poolminer.cpp:246: error: ‘lock’ was not declared in this scope
main_poolminer.cpp:255: error: ‘boost’ has not been declared
main_poolminer.cpp:255: error: expected `;' before ‘io_service’
main_poolminer.cpp:256: error: ‘boost’ has not been declared
main_poolminer.cpp:256: error: expected `;' before ‘resolver’
main_poolminer.cpp:257: error: ‘boost’ has not been declared
main_poolminer.cpp:257: error: expected `;' before ‘query’
main_poolminer.cpp:259: error: ‘boost’ has not been declared
main_poolminer.cpp:259: error: expected `;' before ‘endpoint’
main_poolminer.cpp:260: error: ‘boost’ has not been declared
main_poolminer.cpp:260: error: expected `;' before ‘end’
main_poolminer.cpp:261: error: ‘boost’ has not been declared
main_poolminer.cpp:261: error: expected `;' before ‘nd_option’
main_poolminer.cpp:262: error: ‘boost’ has not been declared
main_poolminer.cpp:262: error: expected `;' before ‘ka_option’
main_poolminer.cpp:265: error: ‘endpoint’ was not declared in this scope
main_poolminer.cpp:265: error: ‘resolver’ was not declared in this scope
main_poolminer.cpp:265: error: ‘query’ was not declared in this scope
main_poolminer.cpp:266: error: ‘boost’ has not been declared
main_poolminer.cpp:266: error: ‘boost’ has not been declared
main_poolminer.cpp:266: error: ‘socket’ was not declared in this scope
main_poolminer.cpp:267: error: ‘boost’ has not been declared
main_poolminer.cpp:267: error: expected `;' before ‘error_socket’
main_poolminer.cpp:268: error: ‘error_socket’ was not declared in this scope
main_poolminer.cpp:268: error: ‘end’ was not declared in this scope
main_poolminer.cpp:271: error: expected type-specifier before ‘boost’
main_poolminer.cpp:271: error: expected `)' before ‘boost’
main_poolminer.cpp:272: error: ‘boost’ has not been declared
main_poolminer.cpp:272: error: expected `;' before ‘tcp_ep’
main_poolminer.cpp:273: error: ‘tcp_ep’ was not declared in this scope
main_poolminer.cpp:276: error: ‘nd_option’ was not declared in this scope
main_poolminer.cpp:277: error: ‘ka_option’ was not declared in this scope
main_poolminer.cpp:279: error: ‘error_socket’ was not declared in this scope
main_poolminer.cpp:281: error: ‘boost’ has not been declared
main_poolminer.cpp:281: error: ‘boost’ has not been declared
main_poolminer.cpp:284: error: ‘t_start’ was not declared in this scope
main_poolminer.cpp:284: error: ‘boost’ has not been declared
main_poolminer.cpp:305: error: ‘boost’ has not been declared
main_poolminer.cpp:305: error: expected `;' before ‘error’
main_poolminer.cpp:306: error: ‘boost’ has not been declared
main_poolminer.cpp:306: error: ‘error’ was not declared in this scope
main_poolminer.cpp:312: error: ‘socket_to_server’ was not declared in this scope
main_poolminer.cpp:320: error: ‘boost’ has not been declared
main_poolminer.cpp:320: error: expected `;' before ‘error’
main_poolminer.cpp:321: error: ‘boost’ has not been declared
main_poolminer.cpp:321: error: ‘boost’ has not been declared
main_poolminer.cpp:321: error: ‘boost’ has not been declared
main_poolminer.cpp:321: error: ‘error’ was not declared in this scope
main_poolminer.cpp:322: error: ‘boost’ has not been declared
main_poolminer.cpp:337: error: ‘boost’ has not been declared
main_poolminer.cpp:337: error: expected `;' before ‘error’
main_poolminer.cpp:338: error: ‘boost’ has not been declared
main_poolminer.cpp:338: error: ‘boost’ has not been declared
main_poolminer.cpp:338: error: ‘boost’ has not been declared
main_poolminer.cpp:338: error: ‘error’ was not declared in this scope
main_poolminer.cpp:339: error: ‘boost’ has not been declared
main_poolminer.cpp:359: error: ‘boost’ has not been declared
main_poolminer.cpp:359: error: expected `;' before ‘error’
main_poolminer.cpp:360: error: ‘boost’ has not been declared
main_poolminer.cpp:360: error: ‘boost’ has not been declared
main_poolminer.cpp:360: error: ‘boost’ has not been declared
main_poolminer.cpp:360: error: ‘error’ was not declared in this scope
main_poolminer.cpp:361: error: ‘boost’ has not been declared
main_poolminer.cpp:406: error: ‘boost’ has not been declared
main_poolminer.cpp:406: error: ‘boost’ has not been declared
main_poolminer.cpp: In member function ‘virtual void CMasterThread::wait_for_master()’:
main_poolminer.cpp:413: error: ‘boost’ has not been declared
main_poolminer.cpp:413: error: ‘boost’ has not been declared
main_poolminer.cpp:413: error: ‘_mutex_master’ was not declared in this scope
main_poolminer.cpp:413: error: ‘lock’ was not declared in this scope
main_poolminer.cpp: In member function ‘void CMasterThread::wait_for_workers()’:
main_poolminer.cpp:423: error: ‘boost’ has not been declared
main_poolminer.cpp:423: error: ‘boost’ has not been declared
main_poolminer.cpp:423: error: ‘_mutex_working’ was not declared in this scope
main_poolminer.cpp:423: error: ‘lock’ was not declared in this scope
main_poolminer.cpp: In member function ‘void CMasterThread::stats_running()’:
main_poolminer.cpp:435: error: ‘setprecision’ is not a member of ‘std’
main_poolminer.cpp:436: error: ‘boost’ has not been declared
main_poolminer.cpp:436: error: expected `;' before ‘t_end’
main_poolminer.cpp:447: error: ‘t_end’ was not declared in this scope
main_poolminer.cpp:448: error: ‘t_start’ was not declared in this scope
main_poolminer.cpp: In function ‘void exit_handler()’:
main_poolminer.cpp:470: error: ‘socket_to_server’ was not declared in this scope
main_poolminer.cpp: In function ‘int main(int, char**)’:
main_poolminer.cpp:598: error: ‘t_start’ was not declared in this scope
main_poolminer.cpp:598: error: ‘boost’ has not been declared
main_poolminer.cpp:612: error: ‘socket_to_server’ was not declared in this scope
main_poolminer.cpp:636: error: aggregate ‘std::stringstream ss’ has incomplete type and cannot be defined
main_poolminer.cpp:637: error: ‘setw’ is not a member of ‘std’
main_poolminer.cpp:637: error: ‘setfill’ is not a member of ‘std’
make: *** [obj/main_poolminer.o] Error 1
Any advice?
-
make -f makefile.unix
In file included from main_poolminer.cpp:15:
main_poolminer.hpp:6:28: error: boost/thread.hpp: No such file or directory
Depending on your flavor of linux (is it linux?):
apt-get install libboost-all-dev
or
yum install boost-devel
-
Keep watching on that, seems CPU mining will be banished from PTS now...
-
Just for reference on mining speeds. My GTX 570 is running at 760c/m and my GTX 560 Ti running at 530 c/m.
-
Yep, linux, I got the the missing package, now this:
make -f makefile.unix
g++ -c -O2 -pthread -DBOOST_SPIRIT_THREADSAFE -D_FILE_OFFSET_BITS=64 -I/root/cudapts/cudapts/src -I/root/cudapts/cudapts/src/obj -fno-stack-protector -fstack-protector-all -Wstack-protector -D_FORTIFY_SOURCE=2 -o obj/main_poolminer.o main_poolminer.cpp
g++ -Wl,-z,relro -Wl,-z,now -o cudapts obj/cpuid.o obj/sha512_avx.o obj/sha512_sse4.o obj/sha512.o obj/sph_sha2.o obj/sph_sha2big.o obj/gpuhash.o obj/main_poolminer.o -L/usr/local/cuda-5.5/lib64 -Wl,-Bdynamic -l boost_system -l boost_filesystem -l boost_program_options -l boost_thread -l boost_chrono -Wl,-Bdynamic -l z -l dl -l cudart -l pthread
/usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: cannot find -lboost_chrono
/usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: cannot find -lz
collect2: ld returned 1 exit status
make: *** [cudapts] Error 1
-
Keep watching on that, seems CPU mining will be banished from PTS now...
Not yet. My miner isn't that fast yet. I can probably speed it up, but the CPU miners can be improved also. It'll be a very fun race for a while. The GPUs will win in the long run, though. But, for example, I have a Haswell desktop (i7-4770) with a GTX 560Ti that mines CPU at the same rate as GPU. It's just that now it can do both at the same time. :) And the GPU draws about 15W while doing the mining instead of the 80W that the CPU draws.
-Dave
-
Yep, linux, I got the the missing package, now this:
make -f makefile.unix
g++ -c -O2 -pthread -DBOOST_SPIRIT_THREADSAFE -D_FILE_OFFSET_BITS=64 -I/root/cudapts/cudapts/src -I/root/cudapts/cudapts/src/obj -fno-stack-protector -fstack-protector-all -Wstack-protector -D_FORTIFY_SOURCE=2 -o obj/main_poolminer.o main_poolminer.cpp
g++ -Wl,-z,relro -Wl,-z,now -o cudapts obj/cpuid.o obj/sha512_avx.o obj/sha512_sse4.o obj/sha512.o obj/sph_sha2.o obj/sph_sha2big.o obj/gpuhash.o obj/main_poolminer.o -L/usr/local/cuda-5.5/lib64 -Wl,-Bdynamic -l boost_system -l boost_filesystem -l boost_program_options -l boost_thread -l boost_chrono -Wl,-Bdynamic -l z -l dl -l cudart -l pthread
/usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: cannot find -lboost_chrono
/usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: cannot find -lz
collect2: ld returned 1 exit status
make: *** [cudapts] Error 1
Which linux? if Ubuntu:
apt-get install
libboost-chrono1.48-dev libboost1.48-doc libboost-date-time1.48-dev libboost-filesystem1.48-dev libboost-graph-parallel1.48-dev libboost-graph1.48-dev libboost-iostreams1.48-dev libboost-locale1.48-dev libboost-math1.48-dev libboost-mpi1.48-dev libboost-program-options1.48-dev libboost-python1.48-dev libboost-random1.48-dev libboost-regex1.48-dev libboost-serialization1.48-dev libboost-signals1.48-dev libboost-system1.48-dev libboost-test1.48-dev libboost-thread1.48-dev libboost-timer1.48-dev libboost-wave1.48-dev xsltproc doxygen default-jdk
and also
apt-get install libz-dev
-
After installing yasm i am getting the following error in Ubuntu 12.04
balu@jagali:~/cudapts$ make -f makefile.unix
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
nvcc -O3 -c -arch=sm_35 -o obj/gpuhash.o gpuhash.cu
make: nvcc: Command not found
make: *** [obj/gpuhash.o] Error 127
-
After installing yasm i am getting the following error in Ubuntu 12.04
balu@jagali:~/cudapts$ make -f makefile.unix
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
nvcc -O3 -c -arch=sm_35 -o obj/gpuhash.o gpuhash.cu
make: nvcc: Command not found
make: *** [obj/gpuhash.o] Error 127
Take a look at page 2, theres a description where to get osfinder.sh. As for nvcc, have you installed the cuda toolkit?
-
After installing yasm i am getting the following error in Ubuntu 12.04
balu@jagali:~/cudapts$ make -f makefile.unix
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
make: ./osfinder.sh: Command not found
nvcc -O3 -c -arch=sm_35 -o obj/gpuhash.o gpuhash.cu
make: nvcc: Command not found
make: *** [obj/gpuhash.o] Error 127
2 problems here:
1) Do a git pull again, I added osfinder
2) You need to have the nvcc compiler in your path. It's in path-to-your-cuda-install/bin
-
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
nvcc fatal : Path to libdevice library not specified
make: *** [obj/gpuhash.o] Error 255
Any idea?
-
Someone please compile it on windows for ypool
-
windows 7 64bit binary?
-
None, yet
-
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
nvcc fatal : Path to libdevice library not specified
make: *** [obj/gpuhash.o] Error 255
Any idea?
Urk, wow. I've never seen that compile error. Is your LD_LIBRARY_PATH set to include the cuda libs? Often /usr/local/cuda-5.5/lib64
But I think this is a CUDA install problem. Googling it suggests that the line
NVVMIR_LIBRARY_DIR = /usr/lib/nvidia-cuda-toolkit/libdevice
to /etc/nvcc.profile
fixed it for some people.
But that may not apply to you. Are you using Debian? If so, give that a shot, correcting the path as needed for the location of the cuda toolkit.
-
I have attempted to compile the code. However, I get the following errors under cygwin 64..
make -f makefile.mingw64
yasm -f win64 -D WINABI -o obj/sha512_avx.o_asm intel/sha512_avx.asm
yasm -f win64 -D WINABI -o obj/sha512_sse4.o_asm intel/sha512_sse4.asm
g++ -c -O2 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sha512.o_amd sha512.c
g++ -c -O1 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sph_sha2.o_a md sph_sha2.c
g++ -c -O1 -march=athlon64 -mmmx -msse -msse2 -fpermissive -o obj/sph_sha2big. o_amd sph_sha2big.c
g++ -c -O2 -march=athlon64 -mmmx -msse -msse2 -mthreads -w -Wall -Wextra -Wform at -Wformat-security -Wno-unused-parameter -DWIN32 -D_WINDOWS -DBOOST_THREAD_US E_LIB -DBOOST_SPIRIT_THREADSAFE -I"/home/jng/cudapts-master/cudapts-master" -I" /usr/local/include" -o obj/main_poolminer.o_amd main_poolminer.cpp
main_poolminer.cpp:18:25: fatal error: sys/syscall.h: No such file or directory
#include <sys/syscall.h>
^
compilation terminated.
makefile.mingw64:66: recipe for target 'obj/main_poolminer.o_amd' failed
make: *** [obj/main_poolminer.o_amd] Error 1
i dont see a file in the github for syscall.h...also, how do I find the boost suffix?, I am almost there, so your help is appreciated.
I am unable to build this as well I keep getting errors with CentOS and osfinder.sh..It seems to no be in the git golder so I downloaded another one but it still doesnt work. I get errors just trying to run the scrip by it self ..I also tried the no chrono version. THat seems to go much further but still fails.
Any help and I will make a contribution to your effort. Thanks
1) I've added osfinder.sh to the git repository. Please re-pull.
2) To the person trying on cygwin, I've removed the need for syscall.h in the most recent commit. Go ahead and update and try again -- I'm not able to test on Cygwin, though, so it's likely you'll encounter further errors.
-Dave
-
Got it working in an ubuntu instance but only getting 540 c/m “Kepler” GK104 / obviously I am doing something wrong
-
Got it working in an ubuntu instance but only getting 540 c/m “Kepler” GK104 / obviously I am doing something wrong
No, that's about what you should expect.
Glad it's working.
-
Excellent thanks!
-
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
nvcc fatal : Path to libdevice library not specified
make: *** [obj/gpuhash.o] Error 255
Any idea?
Urk, wow. I've never seen that compile error. Is your LD_LIBRARY_PATH set to include the cuda libs? Often /usr/local/cuda-5.5/lib64
But I think this is a CUDA install problem. Googling it suggests that the line
NVVMIR_LIBRARY_DIR = /usr/lib/nvidia-cuda-toolkit/libdevice
to /etc/nvcc.profile
fixed it for some people.
But that may not apply to you. Are you using Debian? If so, give that a shot, correcting the path as needed for the location of the cuda toolkit.
It's gentoo. I found them in /opt/cuda/*
Now:
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
/usr/lib/gcc/x86_64-pc-linux-gnu/4.7.3/include/stddef.h:214:32: fatal error: crt/device_runtime.h: No such file or directory
compilation terminated.
make: *** [obj/gpuhash.o] Error 1
Keeping you busy :)
-
Did you install in AWS g2.2xlarge?
-
Instaled Ubuntu 12.04.3, libboost, Yasm and CUDA
ERRO
g++ -c -O2 -fpermissive -o obj/cpuid.o cpuid.c
yasm -f elf64 -o obj/sha512_avx.o intel/sha512_avx.asm
yasm -f elf64 -o obj/sha512_sse4.o intel/sha512_sse4.asm
g++ -c -O2 -fpermissive -o obj/sha512.o sha512.c
g++ -c -O1 -fpermissive -o obj/sph_sha2.o sph_sha2.c
g++ -c -O1 -fpermissive -o obj/sph_sha2big.o sph_sha2big.c
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
make: nvcc: Command not found
make: ** [obj/gpuhash.o] Erro 127
-
Instaled Ubuntu 12.04.3, libboost, Yasm and CUDA
ERRO
g++ -c -O2 -fpermissive -o obj/cpuid.o cpuid.c
make: nvcc: Command not found
You don't have nvcc in your path. You'll need to add it to your PATH and the path to the cuda libraries to your LD_LIBRARY_PATH.
if CUDADIR is the path to where you have cuda installed, in bash, do:
export PATH=$PATH:CUDADIR/bin
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:CUDADIR/lib64
-
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
nvcc fatal : Path to libdevice library not specified
make: *** [obj/gpuhash.o] Error 255
Any idea?
Urk, wow. I've never seen that compile error. Is your LD_LIBRARY_PATH set to include the cuda libs? Often /usr/local/cuda-5.5/lib64
But I think this is a CUDA install problem. Googling it suggests that the line
NVVMIR_LIBRARY_DIR = /usr/lib/nvidia-cuda-toolkit/libdevice
to /etc/nvcc.profile
fixed it for some people.
But that may not apply to you. Are you using Debian? If so, give that a shot, correcting the path as needed for the location of the cuda toolkit.
It's gentoo. I found them in /opt/cuda/*
Now:
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
/usr/lib/gcc/x86_64-pc-linux-gnu/4.7.3/include/stddef.h:214:32: fatal error: crt/device_runtime.h: No such file or directory
compilation terminated.
make: *** [obj/gpuhash.o] Error 1
Keeping you busy :)
Oh gentoo. Um.
The sledgehammer might be to try upgrading to gcc 4.8, but I realize that's a royal pain. You're using the latest CUDA 5.5?
-
Hi,
If you want to check it out in AWS try ami-c4bfd9f4 in Oregon.
sudo nano -w /etc/rc.local
to change start up.
/bin/screen -dmS miner2 bash -c 'until /home/ubuntu/cudapts/src/cudapts <youraddress> 0, 1 && false; do echo "Error, restarting...";done'
this points at beeer as per hardcoded in code
and change yam.cfg in /home/ubuntu for your ypool cpu mining
on start sudo su to see your miners
screen -r miner1
screen -r miner2
This is enough for proof of work :)
If you like this please buy me a beer
PhtFxxNWAHgwzgwAR8vPGJA9soC5qX4SkS
-
Dude, what software did you create there! I already fixed the problem above, but now tell me:
sh: cicc: command not found
Edit: fixed that already, now tell me this:
$ make -f makefile.unix
nvcc -O3 -c -arch=sm_30 -o obj/gpuhash.o gpuhash.cu
cicc: error while loading shared libraries: libnvvm.so.2: wrong ELF class: ELFCLASS32
make: *** [obj/gpuhash.o] Error 127
>:( :D ;D
-
make sure you have the paths set
export PATH=/usr/local/cuda-5.5/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-5.5/lib64:$LD_LIBRARY_PATH
-
Ok i added the 32 bit libs on 64bit arch
working now, thanks!
-
I'm facing this error:
using SSE4
Segmentation fault (core dumped)
my GPU is Quadro 5000 and CPU is Intel(R) Xeon(R) CPU X5650 @ 2.67GHz
Cuda 5.5
-
Good Work .
I would like to suggest you to build a gpu Miner and take about 1% donation (like yvg1900) with every Worker . this would be good for you
We are waiting For Your Release :)
-
I'm facing this error:
using SSE4
Segmentation fault (core dumped)
my GPU is Quadro 5000 and CPU is Intel(R) Xeon(R) CPU X5650 @ 2.67GHz
Cuda 5.5
On unix, change the makefile.unix. Look for where it says sm_30 and change it to sm_20
(for the GPU architecture in the nvcc compilation line)
-Dave
-
Good Work .
I would like to suggest you to build a gpu Miner and take about 1% donation (like yvg1900) with every Worker . this would be good for you
We are waiting For Your Release :)
I agree with you that that's the right way to go. Unfortunately, I don't have the time to do it.
There's a *lot* of support and build-support that needs to be done for a working miner, and GPUs make it even more painful, as the threads here show. yvv1900's yam miner is impressive not just because of its speed, but because he's made it easy to use and work well on a lot of platforms. That takes a lot of attention to engineering -- quite probably more work to do that part than it is just to make a fast miner. I think yvg deserves every 1% he gets out of his miner and that it's a great business model for supporting the development of the code.
I may ask the author of CudaMiner if he wants to do it together, though. :)
-Dave
-
I'm facing this error:
using SSE4
Segmentation fault (core dumped)
my GPU is Quadro 5000 and CPU is Intel(R) Xeon(R) CPU X5650 @ 2.67GHz
Cuda 5.5
On unix, change the makefile.unix. Look for where it says sm_30 and change it to sm_20
(for the GPU architecture in the nvcc compilation line)
-Dave
It works with 30 as well :), I just did not type 0 (device index) at the end of the start line. Thank you!
-
Nice work! Now we need an OpenCL win64 miner :)
-
The pool is http://ptsweb.beeeeer.org/ ?
with 30 works but it freeze, 20 is right parameter for me.
Quadro 5000 gives around 490.8 c/m
-
The pool is http://ptsweb.beeeeer.org/ ?
Yeah, sorry - I just used the one ptsminer supported, which I used because it built easily on my mac. *grin*
If anyone wants to send me a patch to make it work on ypool or other pools, I'm happy to incorporate it.
-
What arch should be for Quadro 4000 ?
-
What arch should be for Quadro 4000 ?
Check out:
https://developer.nvidia.com/cuda-gpus
Compute capability 2.0 -> arch sm_20
-
DGA,
Around here we don't do "contracts" but operate on the basis of trust and informal/non-binding agreements. My entire trust with the community and valuation of PTS, AGS, and future in the bitcoin space is on the line and I am very well known. So that is more of a guarantee than any signed contract, it is in my best interest to do right by everyone I deal with.
It is nice that it is based on ptsminer so does that mean it works with standard pools?
Post the code and when I hear feedback from those on the forum that it works I will pay you 160 PTS.
Dan
Just wanted to follow up publicly to thank you for the bounty and to note that you were good on your word, as promised. I appreciate it.
-Dave
-
Hello,
I've been 'new' to mining since 2 days . Not really new, tried bitcoin about two years back and didn't see where it was going and I friggin can't remember where the wallet is :) . But now I 've been mining LT , which isn't going to well with everyone switching from BT to LT.. So after a long extensive search and reading almost everything there is , PC seems very promising, not only moneywise but also the ideology. Especially that it was only to be for the CPU so only for home users and people abusing servers.. And then I read this thread. So is this the beginning of the end of PC ?
To DGA congratulations, I am already happy when I can write a .cmd file that seems to work :) . BTW If you would happen to have a downloadable miner for AMD videocards, + install instructions , that would be much appreciated. If this is not the end of PC offcourse..
-
Dual GTX 690s: 2400 c/s at 115W at 10% CPU.
650Ti: 330-340 c/s at 1% CPU.
Macbook Pro with GT 650M: 200 c/s.
Tesla K20c: ~800c/s.
My Xeon give me 200 c/m. And you write c/s. This error? How gpu miner faster cpu?
-
Dual GTX 690s: 2400 c/s at 115W at 10% CPU.
650Ti: 330-340 c/s at 1% CPU.
Macbook Pro with GT 650M: 200 c/s.
Tesla K20c: ~800c/s.
My Xeon give me 200 c/m. And you write c/s. This error? How gpu miner faster cpu?
Gaaaaah, thanks. Typo. c/m. c/m c/m c/m.
I'm too used to thinking in kh/s from the other coins.
-Dave
-
..So I guess that means "no" ?
-
http://aws.amazon.com/ec2/ g2.2xlarge
Profitable? Someone counted?
-
Work!
Ubuntu 12.04.3
GTX 570 - 745 C/M
-
..So I guess that means "no" ?
It means "complicated"
Here's an example: One of my systems that I've tested on is a new, nice Haswell-based machine from Dell.
CPU: i7-4770
GPU: GTX 650Ti
CPU hashing gets somewhere in the 330 c/s range using Yam. Maybe more, maybe less, but in that range.
GPU hashing gets almost exactly the same. I think I can sneak it a little higher in the next release, but for now, they're nearly identical.
The CPU costs $300. Not counting motherboard and memory.
The GPU costs about $140, 1GB of DRAM included.
The CPU draws about 70-80W when running full tilt. The GPU uses about 15-30W when running my code full speed.
You need one motherboard (~$120) and DRAM (~$40) for every CPU or two. You can instead fit 3-6 GPUs into a motherboard, reducing those costs further.
Also, my code doesn't really take advantage of the zoominess of something like the 650Ti. People have reported very good numbers from older cards. The memory bandwidth and number of memory channels is pretty important here, and that's not increasing as radically fast as core counts.
In total, running PTS on a GPU probably costs 1/4th as much as doing it on a CPU. That's not insignificant, but it's also not earth-shattering. I suspect that given the current price of PTS and the hash rates, mining PTS on a CPU will actually remain power-profitable, which is more than you can say for most other options (DOGEcoin on a Haswell CPU isn't too bad, though. *grin*)
As I said initially: This doesn't (yet) change the game as radically for PTS as it did for things like Bitcoin. The amount of memory accesses help level the game to a decent degree. But I'll note that I think there's more room for optimization in the GPU code than the CPU code, so the gap will widen a bit more.
-
Update from about 15 minutes ago: I just committed another change to the repository that eliminates a few more errors when trying to compile under Cygwin. It also boosts performance by a fairly negligible amount, at least on compute_35 devices. I didn't bother benchmarking the difference on other cards, because the gain was only about 20 c/m on a Tesla. But, hopefully, this is one step closer to something that will build happily on Windows.
-
Marked, waiting for the window version...
-
Great work!! Just read some code in github, if we use gpu to cal sha512 in the protoshares_revalidateCollision function, it might give a little bit more performance.
-
The following is unscrupulously copied/pasted from my erred compile attempt on Cygwin. I "boostrapped" and compiled the boost library from source, and I don't know whether I did that properly. Also, I'm reading in some places that cygwin won't compile with Windows' network sockets, and that boosts' sockets (?) will more reliably compile and work? I don't know how to switch things around to make that the case. At least its' compiling .o files, though . . .
http://pastebin.com/Sk0c2VjG (http://pastebin.com/Sk0c2VjG)
[Edit: oops! posted that as "code," heh heh]
Also, I've installed MSVC express (which I'm not actually sure cygwin is using?) and CUDA Toolkit 5.5, and I set the paths for those before invoking the cygwin shell, with the following batch:
REM NOTES:
REM A lot of prerequisites, including installing boost on cygwin; re:
REM http://www.boerngen-schmidt.de/2013/07/installing-boost-1-54-0-c-libraries-in-cygwin/
@echo off
call "C:\Program Files\Microsoft Visual Studio 8\Common7\Tools\vsvars32.bat"
REM start cmd /c "C:\Program Files (x86)\Microsoft Visual Studio 11.0\Common7\IDE\WDExpress.exe"
REM Additional environment variables for UCDA/MSVC/GCC compilation;
REM re: https://bitsharestalk.org/index.php?topic=2114.msg25184#msg25184
REM export PATH=$PATH:CUDADIR/bin
REM export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:CUDADIR/lib64
REM
REM OR, adapted:
PATH=%PATH%;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v5.5\bin
LD_LIBRARY_PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v5.5\bin
REM On 32-bit OS:
REM chdir C:\cygwin\bin
REM On 64-bit OS:
chdir C:\cygwin64\bin
bash --login -i
-
Hi all,
the cudapts run great on my GT 560 at 540 cpm, gentoo AMD64.
Anyway, I can not found my shares on http://ptsweb.beeeeer.org/user/<my pts address>.
Is it really working?
-
Hi all,
the cudapts run great on my GT 560 at 540 cpm, gentoo AMD64.
Anyway, I can not found my shares on http://ptsweb.beeeeer.org/user/<my pts address>.
Is it really working?
Yes the user stats does work. If you only just started mining then you'll have to wait till a block is found by the pool.
-
This code is for NVidia cards.
Do we have opensource gpu miners for radeons yet?
-
Hi all,
the cudapts run great on my GT 560 at 540 cpm, gentoo AMD64.
Anyway, I can not found my shares on http://ptsweb.beeeeer.org/user/<my pts address>.
Is it really working?
Yes the user stats does work. If you only just started mining then you'll have to wait till a block is found by the pool.
got it.
thank you very much.
Update: get my first 0.04848 PTS now.
-
How to compile on windows? I know almost nothing about UNIX.
-
dga
are you expecting that AMD cards will be able to mine this faster once someone ports the code like they are with other gpu mined coins?
-
beeeer.org itself very very bad lucky. 5 hours and no block found.
-
beeeer.org itself very very bad lucky. 5 hours and no block found.
Start more miners! :)
-
beeeer.org itself very very bad lucky. 5 hours and no block found.
Start more miners! :)
beeeer.org is very very poor.
since another cuda miner with xpt protocol released, don't use this any more.
-
Got it to compile in Visual Studio 12 but it isn't working ATM
-
Any help with the compiler errors I listed?
Archit, how did you get it to compile in Visual Studio 12?
dga, how did you discover what code abc123 changed? And I'd really rather the latter open source however the hockeysticks he got that (apparently modded) compile . . .
abc123, if you're reading this, any help on how to get it to compile?
-
dga
are you expecting that AMD cards will be able to mine this faster once someone ports the code like they are with other gpu mined coins?
I haven't made up my mind about that yet. The PTS PoW is a little more interesting on GPU than the previous ones, and there are more design choices involved. The actual work of doing the sha512 hashing *will* be faster on AMD, which gives more opportunities to exploit some of the time/space tradeoffs, but Nvidia cards have a nice memory architecture.
Prediction #1: There's a lot of optimization left for GPU PTS mining. My miner is not going to be the fastest one on the block -- I've already improved its speed by about 20% on my development branch, though it uses more memory. I have a few more optimizations in mind. I plan a second release in two weeks once I've had a chance to iron out the kinks. I would be shocked if people didn't at least double the speed of my miner on all architectures in the longer run.
Prediction #2: The AMD-Nvidia gap will be narrower for PTS than for other coins. I don't know which way it will go, though.
Past that, my crystal ball is murky. *grins*
-Dave
-
Any help with the compiler errors I listed?
Archit, how did you get it to compile in Visual Studio 12?
dga, how did you discover what code abc123 changed? And I'd really rather the latter open source however the hockeysticks he got that (apparently modded) compile . . .
abc123, if you're reading this, any help on how to get it to compile?
I didn't - I'm guessing based upon the performance numbers he posted and the memory requirements. *grins* For all I know, he did something completely different. There are several optimizations possible right now.
-Dave
-
https://www.dropbox.com/s/eszhv53k9kitdqz/CUDAPTS.rar
The entire solution is in this file. It requires the boost library for windows and yasm.
dga Did you see what's wrong with it?
-
Compiled with Cygwin, now I have this error,
obj/sha512_avx.o_asm: file not recognized: File format not recognized
collect2: error: ld returned 1 exit status
makefile.mingw64:119: recipe for target 'ptsminer_amd.exe' failed
make: *** [ptsminer_amd.exe] Error 1
any idea?
-
You need YASM
-
Any thoughts on getting it to compile on Mac OS X? The requirements are straightforward using homebrew (brew install boost yasm and ensuring cuda toolkit is installed and in my PATH) but building fails for me on the -march and -mtune switches in cpuid.c.
$ make -f makefile.osx
llvm-g++ -O3 -I/opt/local/include -march=native -Wall -Wextra -Wformat -Wformat-security -Wno-unused-parameter -DMAC_OSX -DMSG_NOSIGNAL=0 -DBOOST_SPIRIT_THREADSAFE -I"/Users/nhyde/Documents/shared/workspace/c/Sandbox/cudapts/src" -I"/Users/nhyde/Documents/shared/workspace/c/Sandbox/cudapts/src"/obj -I"/opt/local/include" -c -O2 -fpermissive -o obj/cpuid.o cpuid.c
cpuid.c:1: error: bad value (native) for -march= switch
cpuid.c:1: error: bad value (native) for -mtune= switch
make: *** [obj/cpuid.o] Error 1
TIA!
-
I'm trying your solution, archit.
It expects YASM to be at:
K:\Downloads\vsyasm-1.2.0-win64
-- I suggest adjusting the path to be relative, with that yasm folder, I don't know--as a subfolder of the project?
I get the project to load without complaint after downloading this, which I assume is the version of YASM you downloaded:
http://www.tortall.net/projects/yasm/releases/vsyasm-1.2.0-win64.zip
--but to get it to load (I don't know how to change the expected path--amateur, here!) I had to make a "Downloads" subfolder in the project, extract yasm under that subfolder, and then map a K: network drive to my own computer (at \\myCompyName\C$\CudaPTS\Downloads). A ridiculous cluge, but it works.
I assume this is the correct version of Boost for my setup (using Microsoft Visual Studio 2012--which is actually version 11 of Visual Studio (brilliant, isn't it?) --
http://sourceforge.net/projects/boost/files/boost-binaries/1.55.0/boost_1_55_0-msvc-11.0-64.exe/download
-- so I'm obtaining that, and trying to integrate it, and we'll see what happens . . .
-
I was in a hurry and wanted to get over with it as soon as possible :P
The boost version I used was http://jaist.dl.sourceforge.net/project/boost/boost/1.55.0/boost_1_55_0.7z but I then had to compile it so I think that would be better.
Tools -> Options -> Projects and Solutions -> VC++ Project Settings -> Build customization search path
-
Working on windows, well almost
-
Works like a dream on an nVidia GT220 1GB on Debian 7 x64.
Working on windows, well almost
Nice work, dude.
-
Working on windows, well almost
Was able to get my build working by forcing SPHLIB. I have a feeling yasm is not playing nice with the ms compiler; maybe the flags need to altered. With SPH forced I am finding and submitting shares, with sse4 or avx it eventually crashes and never submits any work.
-
Same problem, I talked with DGA and he said that he made some "hardcoded assumptions" and he would fix it and then it should work. For now it's just as fast as CPU
-
Was able to get my build working by forcing SPHLIB.
Please clarify what that means/how you do that? Maybe you could post a .rar archive of your project files?
-
Was able to get my build working by forcing SPHLIB.
Please clarify what that means/how you do that? Maybe you could post a .rar archive of your project files?
It's a two-line hack, but it's probably better to just wait; with SPHLIB I am only getting ~95 c/m on a 650 TI instead of the ~335 under linux, which basically puts efficiency below CPU mining. You'd be better off mining an scrypt coin and exchanging it for PTS at that rate.
If you want/need to test yourself, just change: "use_avxsse4 = true" to "use_avxsse4 = false" and comment out Init_SHA512_avx()
-
Working for me, thanks for the contribution.
-
Got it to compile, and it executes, but I need to actually use it with a card that has compute 3.0 ability :)
It's hard-coded to mine at beeeeer.org, apparently, by the way. I'd alter it to accept url etc. parameters . . .
-
Guys I am pretty sure we are going about building a windows release all wrong. Cuda is basicly impossible to get to work under cygwin. The nvida site is riddled with support request about this with nvidias official response being we aren't ever going to support it. Instead I suggest we take Daves gpuhash kernel code which is relatively self contained and easy to get to compile under visual studio and then merge it in with the source for https://github.com/jh000/jhProtominer here. I have gotten the kernel to compile no problem. If anyone who is more familiar with the protoshares algorithm is willing to point me in the right direction of what function call I need to replace with a call to daves kernel I will happily give it a shot and put up a github link with the code.
--Adam
Pretty sure this is what abc123 did btw . As its the only way I can think to get past the yasm roadblock with Dave's source.
-
changed sm to 13
works on gtx 295 np
gtx295 is dual gpu, so 2x 335c/m
Thanks!
-
Earthbound if you just compiled my files, it won't work
Slothlike - I did it but it's a lot slower than abc123
-
I would love to hear how current gen cards fair with this miner (gtx 760, 770, 780). Am I preaching to choir... or will this miner not work with those cards yet?
-
Slothlike - I did it but it's a lot slower than abc123
Archit did you compile for compute 3.5 and did you change the constant that dave recommended (Copied his post below)?
"I see you found which constants to tune to get a little more speed at the use of a little more memory. :-)
If anyone running Linux wants to try to match his performance numbers, change the constant:
#define NUM_COUNTBITS_POWER 31
in gpuhash.cu from 31 to 32. You'll have to have about 1.2GB of memory on your GPU, but if you have it, you'll get a better c/m rate that should match what the above author posted. I plan a future release that auto-selects this a bit more carefully.
n.b. It's perfectly within the license of the code I released to re-brand and add your own donation to it, but I just want to be clear to the forum that these donations aren't going to me. Obviously, however, it's also a pain to get these things working on Windows."
-
The problem isn't that, it's in the hashing algorithm, maybe YASM didn't compile the asm files properly. This forced me to use SPH which I think is creating a bottleneck
-
Got it mostly working 500 c/m on GTX 560ti on windows and ypool :)
I will release it soon
Donations:
PTS:Pn9b5y2FRpAn4V94gTLjxJ9eSQkeuuU6GP
-
how about efficiency in win compares to Linux ?
-
archit was it problems with yasm? I have been fing around with it all night.
-
Is the cuda-miner independent from the cpu? Would this run on a machine with amd-cpu?
-
Thank You, for your great development!
It compiled fine under ubuntu 12.04. Its also running, but does not get any collision:
********************************************
*** cudapts - Nvidia PTS Pool Miner v0.7 RC2 <experimental>
*** by dga - based on ptsminer.
*** If you like this software, please consider sending tips to:
*** PTS: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
*** BTC: 17sb5mcCnnt4xH3eEkVi6kHvhzQRjPRBtS
*** Your donations will encourage further optimization and development
***
*** press CTRL+C to exit
********************************************
using SSE4
spawning 1 worker thread(s)
[WORKER0] Hello, World!
Initializing. Device has 2118770688 free of 2147155968 total bytes of memory
connecting to 54.201.26.128:1337
[WORKER0] GoGoGo!
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
It is an NVIDIA Quadro 3000M GPU and I edited the makefile to sm_20. With 30 is is the same.
What can cause the problem?
-
Thank You, for your great development!
It compiled fine under ubuntu 12.04. Its also running, but does not get any collision:
********************************************
*** cudapts - Nvidia PTS Pool Miner v0.7 RC2 <experimental>
*** by dga - based on ptsminer.
*** If you like this software, please consider sending tips to:
*** PTS: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
*** BTC: 17sb5mcCnnt4xH3eEkVi6kHvhzQRjPRBtS
*** Your donations will encourage further optimization and development
***
*** press CTRL+C to exit
********************************************
using SSE4
spawning 1 worker thread(s)
[WORKER0] Hello, World!
Initializing. Device has 2118770688 free of 2147155968 total bytes of memory
connecting to 54.201.26.128:1337
[WORKER0] GoGoGo!
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
It is an NVIDIA Quadro 3000M GPU and I edited the makefile to sm_20. With 30 is is the same.
What can cause the problem?
Yours should be sm_21. Try and see.
https://developer.nvidia.com/cuda-gpus
-
Thank you. I looked into that and changed it to 21.
Still the same issue and output.
-
dga, thanks for releasing this! I think it's going to be helpful for MemoryCoin users too - do you have an MMC address? I'd like to make a tip.
-
dga, thanks for releasing this! I think it's going to be helpful for MemoryCoin users too - do you have an MMC address? I'd like to make a tip.
Hmm. MMC looks interesting - I'll have to check it out.
Thanks for the tip offer. Donation addresses:
MMC: MTZGEoyforE2ikMWcu9QipHimZgX12Dgom
PTS: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
BTC: 17sb5mcCnnt4xH3eEkVi6kHvhzQRjPRBtS
-
Thanks! 5000 MMC on its way to you. Keep up the good work.
-
Thanks! 5000 MMC on its way to you. Keep up the good work.
Wow - thank you! That's a very nice incentive to keep pounding on this thing. :)
-
dga the Windows version of your code compiled finally and it works
-
dga the Windows version of your code compiled finally and it works
I saw - congrats! You went by re-basing it on jhprotominer and incorporating the mining core, right?
Any changes to the mining core to make it easier next time? I'm going to release some patches in a week that give a decent speedup, and I'd like to not cause people a huge build headache.
-Dave
-
I have just the original version which compiles and runs too! In gpuhash.cu there was a problem with restrict though, you may fix that
-
changed sm to 13
works on gtx 295 np
gtx295 is dual gpu, so 2x 335c/m
Thanks!
test run was old amd x2 cpu,
i put 2 gtx 295 on single intel E5620, all instances hashing at 345+
1380c/m total
-
I have just the original version which compiles and runs too! In gpuhash.cu there was a problem with restrict though, you may fix that
Cool. Yeah, the windows nvcc doesn't seem to like __restrict__ for some reason. Odd.
-
from 0 to ~560c/m @ AWS EC2 g2.2xlarge:
0.
Set up your AWS instance:
In order to create a spot instance you need to get one on demand instance up, set everything up and create an AMI from it.
GPU instances only accept hvm AMIs (scroll down a bit on the Quick Start AMI selection)
This guide was written for Ubuntu 12.04.
Create a key pair if you dont have one already and download it.
When using putty on windows you need puttygen http://www.chiark.greenend.org.uk/~sgtatham/putty/download.html to create a ppk file and select that ppk file under SSH > Auth in the putty menu
User for EC2 instances is ubuntu (on ubuntu instances)
From linux you connect via
ssh -i your_key_file ubuntu@ip_adressMake sure you are root (sudo -s).
1.
Install cuda drivers and openCL:
apt-get update
apt-get -y upgrade
apt-get -y install gcc
wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1204/x86_64/cuda-repo-ubuntu1204_5.5-0_amd64.deb
sudo dpkg -i cuda-repo-ubuntu1204_5.5-0_amd64.deb
apt-get update
apt-get -y install cuda
export PATH=/usr/local/cuda-5.5/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-5.5/lib64:$LD_LIBRARY_PATH
apt-get -y install opencl-headers python-pip python-dev python-numpy python-mako
wget https://pypi.python.org/packages/source/p/pyopencl/pyopencl-2013.1.tar.gz#md5=c506e4ec5bc56ad85bf005ec40d4783b
tar -vxzf pyopencl-2013.1.tar.gz
cd pyopencl-2013.1
python setup.py install
shutdown -r nowSystem needs to reboot before we continue.
2.
Install required packages for building and build the cudapts miner (dont forget to be root):
apt-get -y install git make g++ build-essential libminiupnpc-dev libboost-all-dev libdb++-dev libgmp-dev libssl-dev yasm
apt-get -y install libboost-chrono1.48-dev libboost1.48-doc libboost-date-time1.48-dev libboost-filesystem1.48-dev libboost-graph-parallel1.48-dev libboost-graph1.48-dev libboost-iostreams1.48-dev libboost-locale1.48-dev libboost-math1.48-dev libboost-mpi1.48-dev libboost-program-options1.48-dev libboost-python1.48-dev libboost-random1.48-dev libboost-regex1.48-dev libboost-serialization1.48-dev libboost-signals1.48-dev libboost-system1.48-dev libboost-test1.48-dev libboost-thread1.48-dev libboost-timer1.48-dev libboost-wave1.48-dev
cd ~
git clone https://github.com/dave-andersen/cudapts
cd cudapts/src/
export PATH=$PATH:/usr/local/cuda/bin
make -f makefile.unix
3.
run the miner:
cd ~
export PATH=/usr/local/cuda-5.5/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-5.5/lib64
cudapts/src/cudapts YOUR_PTS_ADRESS 0
4.
???
5.
Profit
(well maybe a bit too optimistic, but i was pretty excited learning all that)
-
2014-01-01 update:
I've committed some new changes to the repository. Some are cosmetic, but three are important:
1) Memory use on the host side is reduced by about 500MB. This may or may not matter for you.
2) Speed is boosted by 10-20% on a lot of platforms. I have another speed boost patch coming next week once I've made it not horrible, but this one gets a decent chunk of the gains.
3) There's now a developer fee that goes to me. Kinda.
I'm doing an experiment with this code release in the developer fee: It's easy to disable. It's not hidden. But it's also just a list of addresses that share the dev fee equally.
So here's my proposal: If you port this software to another platform or release a binary, don't remove my address. Instead, add yours to the list -- I've tried to make it super easy for you to get your own share. If this works out, I'll continue to release improvements and try to make it even easier for other developers who improve upon the code, because we'll all have a reason to make software that remains open source and which is user-friendly and high performance.
If you think this is horrible, let me know and let's try to find a way to make it work better.
If you're a user who hates the idea of a dev fee, the source is yours and you can delete the addresses listed there and/or add your own.
-Dave
-
from 0 to ~560c/m @ AWS EC2 g2.2xlarge:
5.
Profit
(well maybe a bit too optimistic, but i was pretty excited learning all that)
Just curious, is that specific EC2 around $16 a day? And at current diff this would generate 0.2244 PTS?
(http://mrx.im/pts.php?cpm=560) I'm not knocking what you have provided in the way of guidance, I'm checking my math to realize ROI over upgrading some existing equipment (CPU's) or wait for ATI GPU miner as I have stacks of them left over from the early BTC days.
-
from 0 to ~560c/m @ AWS EC2 g2.2xlarge:
5.
Profit
(well maybe a bit too optimistic, but i was pretty excited learning all that)
Just curious, is that specific EC2 around $16 a day? And at current diff this would generate 0.2244 PTS?
(http://mrx.im/pts.php?cpm=560) I'm not knocking what you have provided in the way of guidance, I'm checking my math to realize ROI over upgrading some existing equipment (CPU's) or wait for ATI GPU miner as I have stacks of them left over from the early BTC days.
It's cheaper than that on the spot market, sometimes.
http://aws.amazon.com/ec2/spot-instances/
For example, right now, most markets do cost $16/day on spot (because of all the damn miners. grin.)
But Northern California is only $0.2/hour = $4.80 per day.
Which, alas, is below being profitable.
But, um, if you grab my most recent code you might find that the profitability threshold on ec2 just went up. I'm getting over 800 c/m on Amazon, but my instances only run when the spot market prices drop below about $0.11/hour -- which they don't seem very likely to do.
Note: Running on spot instances is a little scary. It's VERY easy to typo in the box where you say how much you'll pay per hour. DOUBLE AND TRIPLE CHECK YOUR BIDS. From watching the prices historically, it's very clear that some people missed the decimal point and ended up paying more on the spot market than if they just paid the $16/day for a normal g2.2xlarge instance. (!!!)
You also have to watch the market prices of the things you're mining very carefully.
-Dave
-
from 0 to ~560c/m @ AWS EC2 g2.2xlarge:
5.
Profit
(well maybe a bit too optimistic, but i was pretty excited learning all that)
Just curious, is that specific EC2 around $16 a day? And at current diff this would generate 0.2244 PTS?
(http://mrx.im/pts.php?cpm=560) I'm not knocking what you have provided in the way of guidance, I'm checking my math to realize ROI over upgrading some existing equipment (CPU's) or wait for ATI GPU miner as I have stacks of them left over from the early BTC days.
It's cheaper than that on the spot market, sometimes.
http://aws.amazon.com/ec2/spot-instances/
For example, right now, most markets do cost $16/day on spot (because of all the damn miners. grin.)
But Northern California is only $0.2/hour = $4.80 per day.
Which, alas, is below being profitable.
But, um, if you grab my most recent code you might find that the profitability threshold on ec2 just went up. I'm getting over 800 c/m on Amazon, but my instances only run when the spot market prices drop below about $0.11/hour -- which they don't seem very likely to do.
Note: Running on spot instances is a little scary. It's VERY easy to typo in the box where you say how much you'll pay per hour. DOUBLE AND TRIPLE CHECK YOUR BIDS. From watching the prices historically, it's very clear that some people missed the decimal point and ended up paying more on the spot market than if they just paid the $16/day for a normal g2.2xlarge instance. (!!!)
You also have to watch the market prices of the things you're mining very carefully.
-Dave
Thank you for this guidance, I was not aware of the spot instance option. I will research further and be cautious on the bids. I'll also need to learn how to get the mining software to launch automatically as it sounds like if your bid hits the VM will start automatically.(?) Maybe have it send an email as well so you know it's alive and to keep an eye on it.
-
No that was no guide for profitable mining.
At least not profitable today :D
But it still is the easiest way to buy bitcoin via credit card *cough* :o
-
Performance feedback for Speed optimizations #1
GTX 570: Previously 760+ c/m > Now 790+ c/m
GTX 560 Ti: No performance difference 530+ c/m
-
2014-01-01 update:
I've committed some new changes to the repository. Some are cosmetic, but three are important:
1) Memory use on the host side is reduced by about 500MB. This may or may not matter for you.
2) Speed is boosted by 10-20% on a lot of platforms. I have another speed boost patch coming next week once I've made it not horrible, but this one gets a decent chunk of the gains.
3) There's now a developer fee that goes to me. Kinda.
I'm doing an experiment with this code release in the developer fee: It's easy to disable. It's not hidden. But it's also just a list of addresses that share the dev fee equally.
So here's my proposal: If you port this software to another platform or release a binary, don't remove my address. Instead, add yours to the list -- I've tried to make it super easy for you to get your own share. If this works out, I'll continue to release improvements and try to make it even easier for other developers who improve upon the code, because we'll all have a reason to make software that remains open source and which is user-friendly and high performance.
If you think this is horrible, let me know and let's try to find a way to make it work better.
If you're a user who hates the idea of a dev fee, the source is yours and you can delete the addresses listed there and/or add your own.
-Dave
You mean 10-01-2013 update? ;D
Superb work dga. Since I have a few Nvidia cards lying around, I report some of my test since yesterday.
cudarts version 08-01-2013 (v7 if I not mistaken)
GTX 780 - 1450cpm
GTX 680 - 650 cpm
GTX 580 - 850 cpm (3GB memory)
GTX 580 - 920 cpm (1.5GB memory)
GTX 570 - 750 cpm
GTX 260 - 290 cpm
cudarts version 10-01-2013 (v8)
GTX 780 - 1800cpm
GTX 680 - 950 cpm
GTX 580 - 820 cpm - 930 cpm (3GB memory) (the value varies depends on card manufactures)
GTX 580 - 960 cpm (1.5GB memory)
GTX 570 - 770 cpm
GTX 260 - 240 cpm
Yeah most of cards got very nice bump, but I notice some reduction too. But the most nice thing of V8 is my card running at least 3grad C lower. No change on memory consumption.
There some points I still don't understand:
1. Why GTX680 card not much faster than 580. The GTX680 runs with v7 even slower than GTX580.
2. I tried to compile with sm_35 for GTX780 cards. But it got around 10-15% slower than with sm_30.
Anyway I'm very happy with this. Thanks dga.
-
Tried it with my old 9800GT
it didn't print out any statistics (col/min ?)
and i almost burned the card... (it have a broken fan).. when funny smell started i stoped the cudapts :)
-
GTX 780 - 1800cpm
Jesus Christ!
/me went off shopping!
-
GTX 780 - 1800cpm
Jesus Christ!
/me went off shopping!
grin. Now you're seeing which platforms I'm able to develop for. Gorgeous result.
I have a dual GTX690 rig that's currently doing about 3600 c/m. I like it.
-
Any one who can test my miner as well and see how the speeds compare?
-
Any one who can test my miner as well and see how the speeds compare?
I would like too, but without source I can't compile for linux. I have a quite broad range of nvidia cards lying around, so it's very interesting to test, but my computer is all linux so windows binary is impossible for me :).
-
GTX 780 - 1800cpm
Jesus Christ!
/me went off shopping!
grin. Now you're seeing which platforms I'm able to develop for. Gorgeous result.
I have a dual GTX690 rig that's currently doing about 3600 c/m. I like it.
dga, do you have any idea why GTX680 card is not faster than 580? And there is any improvements of sm3.5 over sm3.0? In my test 3.5 card (GTX780) loses 20% cpm when using sm3.5 over sm3.0.
Thanks in advance if anyone could explain it for me.
-
dga, do you have any idea why GTX680 card is not faster than 580? And there is any improvements of sm3.5 over sm3.0? In my test 3.5 card (GTX780) loses 20% cpm when using sm3.5 over sm3.0.
Thanks in advance if anyone could explain it for me.
Because the biggest bottleneck with my current design is memory latency and the number of memory references that can be in flight at a time. This isn't something that improves nearly as fast with different generations of cards (and sometimes gets worse). Though I'm surprised at your 780 slowing down with sm_35. I'll have to try that.
Lesee. K20c I get 1104 c/m with sm_35. Results are nearly unchanged with sm_30.
The biggest crypto-related improvement in sm_35 is the funnel shifter, but the design of Momentum is less crypto-intensive than a lot of PoW functions because it generates 8 memory references for every one execution of SHA512. This is in part why I don't expect there to be as big an nvidia-AMD gap (except that by being so much faster, the AMD crew can try to exploit space/time tradeoffs in different ways.).
I'm playing around with some very different design options for faster cards, but haven't decided on anything I really like yet.
-
Hey, @Archit, @crz, and anyone else who's been porting the software to other platforms:
Does it cause you increased pain if I include any of the Thrust header files?
You could try, e.g., uncommenting thrust/sort.h in gpuhash.cu and see if it breaks anything. (I'll warn that I've actually had the most problems with it at link time when functions are used, though, not just from using the headers).
I'm trying to get rid of the need for yasm by moving a bit more functionality onto the GPU, and using Thrust will let me be lazy about it. *grin*
-Dave
-
I'm sorry to say it increases pain. YASM is not a big problem I can deal with it
-
Hi, test miner on GTX 660.
arCUDAminer - 470c/m (win build from post Reply #91)
Nice!
-
2014-01-01 update:
I've committed some new changes to the repository. Some are cosmetic, but three are important:
1) Memory use on the host side is reduced by about 500MB. This may or may not matter for you.
2) Speed is boosted by 10-20% on a lot of platforms. I have another speed boost patch coming next week once I've made it not horrible, but this one gets a decent chunk of the gains.
3) There's now a developer fee that goes to me. Kinda.
I'm doing an experiment with this code release in the developer fee: It's easy to disable. It's not hidden. But it's also just a list of addresses that share the dev fee equally.
So here's my proposal: If you port this software to another platform or release a binary, don't remove my address. Instead, add yours to the list -- I've tried to make it super easy for you to get your own share. If this works out, I'll continue to release improvements and try to make it even easier for other developers who improve upon the code, because we'll all have a reason to make software that remains open source and which is user-friendly and high performance.
If you think this is horrible, let me know and let's try to find a way to make it work better.
If you're a user who hates the idea of a dev fee, the source is yours and you can delete the addresses listed there and/or add your own.
-Dave
Sorry I missed this. I think this is a good idea. Maybe :)
-
dga, do you have any idea why GTX680 card is not faster than 580? And there is any improvements of sm3.5 over sm3.0? In my test 3.5 card (GTX780) loses 20% cpm when using sm3.5 over sm3.0.
Thanks in advance if anyone could explain it for me.
Because the biggest bottleneck with my current design is memory latency and the number of memory references that can be in flight at a time. This isn't something that improves nearly as fast with different generations of cards (and sometimes gets worse). Though I'm surprised at your 780 slowing down with sm_35. I'll have to try that.
Lesee. K20c I get 1104 c/m with sm_35. Results are nearly unchanged with sm_30.
The biggest crypto-related improvement in sm_35 is the funnel shifter, but the design of Momentum is less crypto-intensive than a lot of PoW functions because it generates 8 memory references for every one execution of SHA512. This is in part why I don't expect there to be as big an nvidia-AMD gap (except that by being so much faster, the AMD crew can try to exploit space/time tradeoffs in different ways.).
I'm playing around with some very different design options for faster cards, but haven't decided on anything I really like yet.
Someone over 1gh pool did something similar like you, dga.
https://bitsharestalk.org/index.php?topic=1784.0
I just gave it a short test, and so far I can say that this GPU miner has some similar performance with your miner ( around 10% more depends on cards but it uses at least 2 threads of CPU with 100% load). Temperature of cards stay the same but more memory is required (1.2GB).
Since he doesn't public his code I can say for sure but it maybe use your code. And there some reports that it works with AMD cards too.
PTS mining is exploding in a few days.
-
dga, do you have any idea why GTX680 card is not faster than 580? And there is any improvements of sm3.5 over sm3.0? In my test 3.5 card (GTX780) loses 20% cpm when using sm3.5 over sm3.0.
Thanks in advance if anyone could explain it for me.
Because the biggest bottleneck with my current design is memory latency and the number of memory references that can be in flight at a time. This isn't something that improves nearly as fast with different generations of cards (and sometimes gets worse). Though I'm surprised at your 780 slowing down with sm_35. I'll have to try that.
Lesee. K20c I get 1104 c/m with sm_35. Results are nearly unchanged with sm_30.
The biggest crypto-related improvement in sm_35 is the funnel shifter, but the design of Momentum is less crypto-intensive than a lot of PoW functions because it generates 8 memory references for every one execution of SHA512. This is in part why I don't expect there to be as big an nvidia-AMD gap (except that by being so much faster, the AMD crew can try to exploit space/time tradeoffs in different ways.).
I'm playing around with some very different design options for faster cards, but haven't decided on anything I really like yet.
Someone over 1gh pool did something similar like you, dga.
https://bitsharestalk.org/index.php?topic=1784.0
I just gave it a short test, and so far I can say that this GPU miner has some similar performance with your miner ( around 10% more depends on cards but it uses at least 2 threads of CPU with 100% load). Temperature of cards stay the same but more memory is required (1.2GB).
Since he doesn't public his code I can say for sure but it maybe use your code. And there some reports that it works with AMD cards too.
PTS mining is exploding in a few days.
Well, let's not let the AMD folks win, shall we?
Obviously, they can keep cloning my software, but let's let the folks using the open version have some fun for a while longer.
I've committed another optimization to the repository. My GTX690 is now getting about 1850 c/m - not a huge boost, but a little.
The more important thing about this commit is that it removes the dependency upon yasm, which I know was causing some folks a headache trying to compile it on Windows. Let's see if this makes it easier to use the open source version instead of a who-knows-what-you're-getting closed clone. :-)
-Dave
-
Well, let's not let the AMD folks win, shall we?
Obviously, they can keep cloning my software, but let's let the folks using the open version have some fun for a while longer.
I've committed another optimization to the repository. My GTX690 is now getting about 1850 c/m - not a huge boost, but a little.
The more important thing about this commit is that it removes the dependency upon yasm, which I know was causing some folks a headache trying to compile it on Windows. Let's see if this makes it easier to use the open source version instead of a who-knows-what-you're-getting closed clone. :-)
-Dave
And - many thanks to the people who're running the miner with dev donations to me enabled. I know you don't need to, and I appreciate it. That, and the few direct donations I've received, are what's going to keep these improvements flowing.
-Dave
-
with the new version it's look like im down 100c/m per gtx 295
each gpu in the gtx 295 was getting 340+ now it is only getting 290c/m
-
with the new version it's look like im down 100c/m per gtx 295
each gpu in the gtx 295 was getting 340+ now it is only getting 290c/m
+1 to this - my GT210 dropped from 80c/m to 50c/m with the latest update.
-
with the new version it's look like im down 100c/m per gtx 295
each gpu in the gtx 295 was getting 340+ now it is only getting 290c/m
+1 to this - my GT210 dropped from 80c/m to 50c/m with the latest update.
That's no good. :(
Did you both let it run for a while before comparing the #s? It can take an hour or two (particularly with slower cards) for the rate to become steady. I'm going to add another "speed" metric that's a bit more relevant to quick benchmarking, but for now, c/m it is. :)
I'm surprised. The changes I made shouldn't be large, and slower cards, if anything, should see almost no effect.
-
with the new version it's look like im down 100c/m per gtx 295
each gpu in the gtx 295 was getting 340+ now it is only getting 290c/m
+1 to this - my GT210 dropped from 80c/m to 50c/m with the latest update.
That's no good. :(
Did you both let it run for a while before comparing the #s? It can take an hour or two (particularly with slower cards) for the rate to become steady. I'm going to add another "speed" metric that's a bit more relevant to quick benchmarking, but for now, c/m it is. :)
I'm surprised. The changes I made shouldn't be large, and slower cards, if anything, should see almost no effect.
I can confirm too, dga.
I have test at least each card for 1 hour but the maximal rate drops at least 10% in comparison to your previous version. I don't check the memory consumption or temperature.
-
with the new version it's look like im down 100c/m per gtx 295
each gpu in the gtx 295 was getting 340+ now it is only getting 290c/m
+1 to this - my GT210 dropped from 80c/m to 50c/m with the latest update.
That's no good. :(
Did you both let it run for a while before comparing the #s? It can take an hour or two (particularly with slower cards) for the rate to become steady. I'm going to add another "speed" metric that's a bit more relevant to quick benchmarking, but for now, c/m it is. :)
I'm surprised. The changes I made shouldn't be large, and slower cards, if anything, should see almost no effect.
Yessir I did - normally let it run for ~2 hours to stabilize. :(
-
Gronk make dirty hack visual studio windows port compile and run. Gronk post compile and source code here.
Gronk not sure miner actually doing any pool work when run. GPU gears spin-spin, but no output to screen maybe? How make write on screen? Is doing any work? Maybe it work? Please test, advise. Gronk tax disabled for test. Mines at Gronk drink beeeeer mine. Want make mine anywhere. Not do that for now.
Thank!
Miner in Gronk /src/Release cave.
https://www.dropbox.com/s/q9u4r8sy0zsgf7j/cudaPTSwin.7z
-
earthbound,
got missing MSVCP110D.dll
Thx
-
earthbound,
got missing MSVCP110D.dll
Thx
Looking into that, apparently that means I didn't properly build it in "release" mode (so even installing the redisributable probably wouldn't fix it). . .
-
Ok just released cudaPTSwin - Beeeeer - PTS GPU miner for Windows -> https://bitsharestalk.org/index.php?topic=2311.0
Cheers!
-
For those who don't feel like reading through a bunch of source code but find the algorithmic issues interesting, I've posted a little writeup on my blog about the basic problem and solution used in the GPU PTS miner:
http://da-data.blogspot.com/2014/01/gaining-momentum-duplicate-detection-in.html
-Dave
-
The latest build dropped my cm rate from ~1650 to ~1250 on a GTX 780.
-
For those who don't feel like reading through a bunch of source code but find the algorithmic issues interesting, I've posted a little writeup on my blog about the basic problem and solution used in the GPU PTS miner:
http://da-data.blogspot.com/2014/01/gaining-momentum-duplicate-detection-in.html
-Dave
Hi dga.
Your work is fantastic. But I do have a small request for the next build, and I would like to ask if it's possible: Could you implement the interactive mode like cudaminer. Your miner is already top, efficient and in low temperature for the card, but I can't use the computer anymore if I let it runs. Cudaminer has the interactive switch, so if you switch this on, you lose about 5-10% of hash, but you won't notice much about how your computer lags. I would be very happy with this.
Thanks dga.
Regards.
-
hi. How to use cudapts to connect to ypool.net?
-
For those who don't feel like reading through a bunch of source code but find the algorithmic issues interesting, I've posted a little writeup on my blog about the basic problem and solution used in the GPU PTS miner:
http://da-data.blogspot.com/2014/01/gaining-momentum-duplicate-detection-in.html
-Dave
Hi dga.
Your work is fantastic. But I do have a small request for the next build, and I would like to ask if it's possible: Could you implement the interactive mode like cudaminer. Your miner is already top, efficient and in low temperature for the card, but I can't use the computer anymore if I let it runs. Cudaminer has the interactive switch, so if you switch this on, you lose about 5-10% of hash, but you won't notice much about how your computer lags. I would be very happy with this.
Thanks dga.
Regards.
Yup, it's on the TODO.
-
... I suggest we take Daves gpuhash kernel code which is relatively self contained and easy to get to compile under visual studio and then merge it in with the source for https://github.com/jh000/jhProtominer here. I have gotten the kernel to compile no problem. If anyone who is more familiar with the protoshares algorithm is willing to point me in the right direction of what function call I need to replace with a call to daves kernel I will happily give it a shot and put up a github link with the code...
Maybe this post from iruu could point you in the right direction?
https://bitsharestalk.org/index.php?topic=2225.msg26559#msg26559
If you'd like to PM me with a link to your source archive, I could poke around in it . . . my intent is to create something that we'd release open-source (if we get it working).
-
Maybe this post from iruu could point you in the right direction? . . .
Also, I have no idea whether it may help, but you're free to look at the Visual Studio project source files I posted in this archive:
https://www.dropbox.com/s/q9u4r8sy0zsgf7j/cudaPTSwin.7z
This got all of the files to compile, it's just that the miner doesn't seem to properly communicate with a mining pool. With a traffic monitor, you can see traffic sending out from it, but it seems that no traffic is sent back from the pool; it also eventually throws an error which I looked up and which has to do with not connecting to a network.
A hosts file hack which has been used for other miners also works for this (or, it might, if this miner worked :) ), which allows you to redirect a miner hard-coded to one pool to mine at another (of the same coin and protocol). It's hard-coded right now to beeeeer.org, but by using this hack:
https://bitsharestalk.org/index.php?topic=1921.msg22358#msg22358 (https://bitsharestalk.org/index.php?topic=1921.msg22358#msg22358)
You can get it to try connecting to e.g. mmcpool.1gh.com. Er, if that pool uses the same port :/
-
I have a dual GTX690 rig that's currently doing about 3600 c/m. I like it.
Dear dga,
There are 2 GPU on 1 card GTX690 and it' s mean that 1 card GTX690 get 3600 c/m ? Or you use 2 card GTX690 to get 3600 c/m?
-
I have a dual GTX690 rig that's currently doing about 3600 c/m. I like it.
Dear dga,
There are 2 GPU on 1 card GTX690 and it' s mean that 1 card GTX690 get 3600 c/m ? Or you use 2 card GTX690 to get 3600 c/m?
Two cards. One GTX690 gets 900 cpm on each of its chips, for a total of about 1800cpm per card.
-
hi. How to use cudapts to connect to ypool.net?
EDIT: That may not work: does ypool use a different work-sharing protocol?
I detail a hack here:
https://bitsharestalk.org/index.php?topic=2338.msg28809#msg28809
--but I suggest using beeeeer.org, as (last I heard) ypool had vastly more than 51% of all hashing power going for this coin.
-
I compiled cudapts on a Laptop with fx2800M nvidia card,
but the program aborted when I run it:
%./cudapts Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc 0 sse4
********************************************
*** cudapts - Nvidia PTS Pool Miner v0.8 GPU0.2 <experimental>
*** by dga - based on ptsminer.
*** If you like this software, please consider sending tips to:
*** PTS: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
*** BTC: 17sb5mcCnnt4xH3eEkVi6kHvhzQRjPRBtS
*** Your donations will encourage further optimization and development
***
*** press CTRL+C to exit
********************************************
spawning 1 worker thread(s)
[WORKER0] starting
Initializing. Device has 950116096 free of 1073020928 total bytes of memory
[WORKER0] GoGoGo!
terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::system::system_error> >'
what(): Host not found (authoritative)
Aborted (core dumped)
-
I compiled cudapts on a Laptop with fx2800M nvidia card,
but the program aborted when I run it:
%./cudapts Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc 0 sse4
********************************************
*** cudapts - Nvidia PTS Pool Miner v0.8 GPU0.2 <experimental>
*** by dga - based on ptsminer.
*** If you like this software, please consider sending tips to:
*** PTS: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
*** BTC: 17sb5mcCnnt4xH3eEkVi6kHvhzQRjPRBtS
*** Your donations will encourage further optimization and development
***
*** press CTRL+C to exit
********************************************
spawning 1 worker thread(s)
[WORKER0] starting
Initializing. Device has 950116096 free of 1073020928 total bytes of memory
[WORKER0] GoGoGo!
terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::system::system_error> >'
what(): Host not found (authoritative)
Aborted (core dumped)
it turned out to be a network problem, it works after I PINGed beeeer pool,
the speed is around 102-105 cpm:
[MASTER] submitted share -> SHARE
[STATS] 2014-Jan-18 22:28:17 | 105.4 c/m | 0.8 sh/m | VL: 1 (100.0%), RJ: 0 (0.0%), ST: 0 (0.0%)
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[WORKER] collision found: 55070148 <-> 55597815 #314 @ 1390055418 by 0
[MASTER] submitted share -> REJECTED
[STATS] 2014-Jan-18 22:30:06 | 103.0 c/m | 0.7 sh/m | VL: 1 (50.0%), RJ: 1 (50.0%), ST: 0 (0.0%)
[WORKER] collision found: 38377421 <-> 57999410 #368 @ 1390055468 by 0
[MASTER] submitted share -> REJECTED
[STATS] 2014-Jan-18 22:30:40 | 102.3 c/m | 0.8 sh/m | VL: 1 (33.3%), RJ: 2 (66.7%), ST: 0 (0.0%)
-
How 512mb memory support
-
I'd be curious on how's it work here :D
http://www.nvidia.com/object/gpu-test-drive.html
-
can you mine protoshares with AMD GPU?
-
GT630 rev.2 GK208 384 cores, 902/1600Mhz ~110cpm
-
dga, where do I have to look in code to modify it so it can run on 512 MB cards?
-
dga, where do I have to look in code to modify it so it can run on 512 MB cards?
You'd have to substantially change it - its basic step is that it computes all of the hashes (which require 512MB) and then does some work on them. This isn't a tuning parameter, it's fundamental to the current design.
-
dga, will you further improve it?
-
It crashes on my system.
Ubuntu 12.04.03 x86.
GTX 470
nvidia driver 304
cuda toolkit 5.0
user@host:~/bitcoin/build/cudapts-src/src$ gdb ./cudapts
Reading symbols from /home/k/bitcoin/build/cudapts-src/src/cudapts...(no debugging symbols found)...done.
(gdb) run PuhdpvPc1rpe9yfYeGKJ5r8tqLaAPoERMm 0
Starting program: /home/k/bitcoin/build/cudapts-src/src/cudapts PuhdpvPc1rpe9yfYeGKJ5r8tqLaAPoERMm 0
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1".
********************************************
*** cudapts - Nvidia PTS Pool Miner v0.8 GPU0.2 <experimental>
*** by dga - based on ptsminer.
*** If you like this software, please consider sending tips to:
*** PTS: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
*** BTC: 17sb5mcCnnt4xH3eEkVi6kHvhzQRjPRBtS
*** Your donations will encourage further optimization and development
***
*** press CTRL+C to exit
********************************************
spawning 1 worker thread(s)
[New Thread 0xb7c3bb40 (LWP 24369)]
[WORKER0] starting
[New Thread 0xb67dbb40 (LWP 24370)]
Initializing. Device has 14385152 free of 1341718528 total bytes of memory
Could not malloc dev_data (2)
[WORKER0] GoGoGo!
connecting to 54.201.26.128:1337
Mining for approx 20 seconds to support further development
Payments to: Pr8cnhz5eDsUegBZD4VZmGDARcKaozWbBc
[MASTER] work received - sharetarget: 00ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[WORKER] collision found: 1600286575 <-> 1600286575 #86764 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #86764 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #86962 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #86962 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #87008 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #87008 @ 1391584732 by 0
[WORKER] collision found: 59 <-> 41 #87996 @ 1391584732 by 0
[WORKER] collision found: 1718960741 <-> 1919439973 #89792 @ 1391584732 by 0
[WORKER] collision found: 3073675520 <-> 3073673448 #91980 @ 1391584732 by 0
[WORKER] collision found: 4201475 <-> 4201475 #93708 @ 1391584732 by 0
[WORKER] collision found: 4201475 <-> 4201475 #93708 @ 1391584732 by 0
[WORKER] collision found: 4201475 <-> 4201475 #94564 @ 1391584732 by 0
[WORKER] collision found: 4201475 <-> 4201475 #94564 @ 1391584732 by 0
[WORKER] collision found: 210885635 <-> 210885635 #101250 @ 1391584732 by 0
[WORKER] collision found: 210885635 <-> 210885635 #101250 @ 1391584732 by 0
[WORKER] collision found: 268442624 <-> 268442624 #102428 @ 1391584732 by 0
[WORKER] collision found: 268442624 <-> 268442624 #102428 @ 1391584732 by 0
[WORKER] collision found: 3073797752 <-> 3073797592 #104010 @ 1391584732 by 0
[WORKER] collision found: 1970495300 <-> 1970495297 #107318 @ 1391584732 by 0
[WORKER] collision found: 262143 <-> 196607 #107692 @ 1391584732 by 0
[WORKER] collision found: 32 <-> 120 #112028 @ 1391584732 by 0
[WORKER] collision found: 3073902016 <-> 4294967295 #114450 @ 1391584732 by 0
[WORKER] collision found: 3073902016 <-> 4294967295 #114700 @ 1391584732 by 0
[WORKER] collision found: 1768710495 <-> 0 #115340 @ 1391584732 by 0
[WORKER] collision found: 184550144 <-> 185602560 #125122 @ 1391584732 by 0
[WORKER] collision found: 134224931 <-> 134224931 #130044 @ 1391584732 by 0
[WORKER] collision found: 134224931 <-> 134224931 #130044 @ 1391584732 by 0
[WORKER] collision found: 81920 <-> 82432 #134816 @ 1391584732 by 0
[WORKER] collision found: 3739081947 <-> 3739081947 #139488 @ 1391584732 by 0
[WORKER] collision found: 3739081947 <-> 3739081947 #139488 @ 1391584732 by 0
[WORKER] collision found: 3049009188 <-> 3049009188 #141254 @ 1391584732 by 0
[WORKER] collision found: 3049009188 <-> 3049009188 #141254 @ 1391584732 by 0
[WORKER] collision found: 3049013860 <-> 3049013860 #141294 @ 1391584732 by 0
[WORKER] collision found: 3049013860 <-> 3049013860 #141294 @ 1391584732 by 0
[WORKER] collision found: 1852404853 <-> 1852404853 #142318 @ 1391584732 by 0
[WORKER] collision found: 1852404853 <-> 1852404853 #142318 @ 1391584732 by 0
[WORKER] collision found: 3066482000 <-> 3066482000 #144614 @ 1391584732 by 0
[WORKER] collision found: 3066482000 <-> 3066482000 #144614 @ 1391584732 by 0
[WORKER] collision found: 4201475 <-> 4201475 #145772 @ 1391584732 by 0
[WORKER] collision found: 4201475 <-> 4201475 #145772 @ 1391584732 by 0
[WORKER] collision found: 35940 <-> 35940 #146050 @ 1391584732 by 0
[WORKER] collision found: 35940 <-> 35940 #146050 @ 1391584732 by 0
[WORKER] collision found: 130877512 <-> 130877512 #146652 @ 1391584732 by 0
[WORKER] collision found: 130877512 <-> 130877512 #146652 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #147004 @ 1391584732 by 0
[WORKER] collision found: 1600286575 <-> 1600286575 #147004 @ 1391584732 by 0
[WORKER] collision found: 3073755672 <-> 3073755672 #147962 @ 1391584732 by 0
[WORKER] collision found: 3073755672 <-> 3073755672 #147962 @ 1391584732 by 0
[WORKER] collision found: 3491847427 <-> 3491847427 #150440 @ 1391584732 by 0
[WORKER] collision found: 3491847427 <-> 3491847427 #150440 @ 1391584732 by 0
[WORKER] collision found: 2684427267 <-> 2684427267 #150476 @ 1391584732 by 0
[WORKER] collision found: 2684427267 <-> 2684427267 #150476 @ 1391584732 by 0
[WORKER] collision found: 9437345 <-> 9437345 #150648 @ 1391584732 by 0
[WORKER] collision found: 9437345 <-> 9437345 #150648 @ 1391584732 by 0
[WORKER] collision found: 1768697142 <-> 1768697142 #150936 @ 1391584732 by 0
[WORKER] collision found: 1768697142 <-> 1768697142 #150936 @ 1391584732 by 0
[WORKER] collision found: 20640 <-> 20640 #150982 @ 1391584732 by 0
[WORKER] collision found: 20640 <-> 20640 #150982 @ 1391584732 by 0
[WORKER] collision found: 3071296615 <-> 3071296615 #151188 @ 1391584732 by 0
[WORKER] collision found: 3071296615 <-> 3071296615 #151188 @ 1391584732 by 0
[WORKER] collision found: 436207674 <-> 436207674 #151410 @ 1391584732 by 0
[WORKER] collision found: 436207674 <-> 436207674 #151410 @ 1391584732 by 0
[WORKER] collision found: 32388 <-> 32388 #151664 @ 1391584732 by 0
[WORKER] collision found: 32388 <-> 32388 #151664 @ 1391584732 by 0
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb7c3bb40 (LWP 24369)]
0x0806c5f4 in void protoshares_process_512<2097152, -2097152, 21, (SHAMODE)0>(blockHeader_t*, CBlockProvider*, unsigned int, GPUHasher*, unsigned long long*) ()
(gdb) bt
#0 0x0806c5f4 in void protoshares_process_512<2097152, -2097152, 21, (SHAMODE)0>(blockHeader_t*, CBlockProvider*, unsigned int, GPUHasher*, unsigned long long*) ()
#1 0x0806cc1b in void CWorkerThread::mineloop<2097152, -2097152, 21, (SHAMODE)0>() ()
#2 0x0806d65d in CWorkerThread::run() ()
#3 0xb7faa1cc in ?? () from /usr/lib/libboost_thread.so.1.48.0
#4 0xb7f2dd4c in start_thread () from /lib/i386-linux-gnu/libpthread.so.0
#5 0xb7d69bae in clone () from /lib/i386-linux-gnu/libc.so.6
(gdb)