BitShares Forum
Other => Graveyard => BitShares PTS => Topic started by: MisO69 on November 13, 2013, 03:15:10 pm
-
Post your CPU model and mining speed with jhProtominer 0.1c. Lets get started on a hardware list. I'm curious to see the differences between the linux versions and windows.
These are Windows X64 stats:
UPDATED VERSIONS -=> https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz
jhProtominer optimized by yvg1900. Working link: https://mega.co.nz/#F!h0tkXSxZ!f62uoUXogkxQmP2xO8Ib-g
I have been doing some testing with this optimized miner for ypool. Here are my results. I have tested other platforms on each cpu. For example, corei7avx, generic and core2 have all been tested on my Xeons. I had the best results from the core2 versions so I only posted their stats. Optimized miner stats are bolded.
CPU Model/RAM - Collisions per minute - Miner settings
Core i5 3570 @ 3.4Ghz/4Gb 125 4 threads –m512
Core i5 3570 @ 3.4Ghz/4Gb 163 4 threads corei7avx-512m
Core i5 3570 @ 3.4Ghz/4Gb 114 4 threads –m256
Core i7 870 @ 3Ghz/4Gb 88 4 threads –m512
Core i7 870 @ 3Ghz/4Gb 115 4 threads corei7-512m
Core i7 870 @ 3Ghz/4Gb 101 8 threads –m256
Core i7 870 @ 3Ghz/4Gb 142 8 threads corei7-256m
AMD FX8350 @ 4Ghz/8Gb 190 8 threads –m512
AMD FX8350 @ 4Ghz/8Gb 246 8 threads bulldozerv1-512m
AMD FX8350 @ 4Ghz/8Gb 182 8 threads –m256
Dual Xeon X5660 @ 2.8Ghz/24Gb 246 20 threads –m256
Dual Xeon X5660 @ 2.8Ghz/24Gb 330 20 threads core2-256m
Dual Xeon E5430 @ 2.6Ghz/4Gb 73 8 threads –m256
Dual Xeon E5-2450 @ 2.1Ghz/24Gb 325 32 threads –m512
Dual Xeon E5-2450 @ 2.1Ghz/24Gb 430 32 threads core2-512m
Dual Xeon E5-2450 @ 2.1Ghz/24Gb 283 32 threads –m256
Core2 Q9400 @ 2.6Ghz/4Gb 62 4 threads –m512
Core2 Q9400 @ 2.6Ghz/4Gb 75 4 threads core2-512m
Core2 Q9400 @ 2.6Ghz/2Gb 58 4 threads –m256
Core2 Q9400 @ 2.6Ghz/2Gb 67 4 threads core2-256m
Core2 Q9400 @ 2.6Ghz/2Gb 35 4 threads –m128
AMD Phenom II x4 965 @ 3.4Ghz 76 4 threads -m512
-
Server 2012
Core i7 4770K @ 4.5Ghz/16Gb 206 8 threads –m512
Xeon E5-2630 @ 2.3Ghz/32Gb 115 12 threads –m512
Server 2008 R2
Xeon E5520 @ 2.27Ghz/12Gb 70 8 threads –m256
-
Amd Athlon x640 : 60
-
using ptsminer v0.5 (http://bitsharestalk.org/index.php?topic=234.msg5632#msg5632) based on jh00's work
Intel Core i5 3570k @ 3.4 GHz/8Gb 99.5 3 threads "m512"
-
Amazon AWS cc2.8xlarge, 32 Threads, 279 cpm jhProtominer -m512
Amazon AWS cc2.8xlarge, 32 Threads, 309 cpm ptsminer 0.5 based on jh's
-
Amazon AWS cc2.8xlarge, 32 Threads, 279 cpm jhProtominer -m512
Amazon AWS cc2.8xlarge, 32 Threads, 309 cpm ptsminer 0.5 based on jh's
Have you calculated it's positive ROI to run it on Amazon?
-
i7 3630QM @ 2.4 GHz - 110 collisions/min - jhProtominer -t 8 -m512
AMD Phenom II x4 965 Black edition @ 3.4 GHz - 70 collisions/min - jhProtominer -t 4 -m512
-
Intel Core i5 4570 @ 3.2 GHz/8Gb 116 4 threads m512
-
Amazon AWS cc2.8xlarge, 32 Threads, 279 cpm jhProtominer -m512
Amazon AWS cc2.8xlarge, 32 Threads, 309 cpm ptsminer 0.5 based on jh's
Have you calculated it's positive ROI to run it on Amazon?
Yes if you use spot instances for 0.488$/h.
-
AMD Opteron (tm) Processor 4171 HE 2.09GHz (2 processors) 90 8 threads –m512
-
Vista64
Xeon L5639 @ 3.2Ghz/6Gb 165hpm 10 threads -m512
I need more then 6gb ram to run 11-12 threads.
-
Vista64
Xeon L5639 @ 3.2Ghz/6Gb 162hpm 10 threads -m512
I need more then 6gb ram to run 11-12 threads.
L5639 overclocked? I remember it is 2.13GHz.
Sent from my iPhone using Tapatalk (http://tapatalk.com/m?id=1)
-
Vista64
Xeon L5639 @ 3.2Ghz/6Gb 162hpm 10 threads -m512
I need more then 6gb ram to run 11-12 threads.
L5639 overclocked? I remember it is 2.13GHz.
Sent from my iPhone using Tapatalk (http://tapatalk.com/m?id=1)
Yes overclocked. Drops down to x16 multiplier when all 6 cores are loaded however :(
-
Amazon AWS cc2.8xlarge, 32 Threads, 279 cpm jhProtominer -m512
Amazon AWS cc2.8xlarge, 32 Threads, 309 cpm ptsminer 0.5 based on jh's
cc2.8xlarge, dual Xeon E5-2670, 847 cpm jhProtominer-yvg1900 32Threads * 1024MB
Update: updated miner, ~900cpm now.
Update: updated miner again , ~955cpm now.
Update: updated miner again , ~985cpm now.
BTW: another machine E3-1230 V3, 8Threads * 1024MB, 336 cpm
-
Vista64
Xeon L5639 @ 3.2Ghz/6Gb 165hpm 10 threads -m512
I need more then 6gb ram to run 11-12 threads.
Have you tried running it with -m256 and using all 12 threads?
-
Collisions are misleading in ypool.
Take in consideration for PPS rate and fees.
ypool has MASSIVE fees.
Experiment yourself.
-
These are with Tydus' 'large memory' mod on Windows 7 x64 SP1
Core i7-920 (4 cores, no HT, 3.07GHz) ~100col/min -t 4 -m 1024
Core i7-920 (4 cores, no HT, 3.07GHz) ~94col/min -t 4 -m 2048
Core i7-920 (4 cores, no HT, 3.07GHz) ~91col/min -t 4 -m 4096
Currently running it at -m4096 even though it's slightly lower collisions to see if anything else is really getting a bonus from the large memory usage, but the 'sweet spot' for highest col/min is 1024/thread for me. Running at -m512 ended up only netting me 75-80col/min.
Also to note that using the same hardware as above i was getting ~90col/min on ptsminer v0.5a x64.
Anyone know what the larger memory footprint is meant to do? I assume keep more of the existing hashes in memory to provide references for collisions but I haven't actually seen it 'work that way' from what I've read and seen.
-
These are with Tydus' 'large memory' mod on Windows 7 x64 SP1
Core i7-920 (4 cores, no HT, 3.07GHz) ~100col/min -t 4 -m 1024
Core i7-920 (4 cores, no HT, 3.07GHz) ~94col/min -t 4 -m 2048
Core i7-920 (4 cores, no HT, 3.07GHz) ~91col/min -t 4 -m 4096
Currently running it at -m4096 even though it's slightly lower collisions to see if anything else is really getting a bonus from the large memory usage, but the 'sweet spot' for highest col/min is 1024/thread for me. Running at -m512 ended up only netting me 75-80col/min.
Also to note that using the same hardware as above i was getting ~90col/min on ptsminer v0.5a x64.
Anyone know what the larger memory footprint is meant to do? I assume keep more of the existing hashes in memory to provide references for collisions but I haven't actually seen it 'work that way' from what I've read and seen.
Do you have a link for Tydus large memory miner for Windows x64? I looked around and only could find the source on github. I would like to try it on my 32 thread 24gig dual xeon monstrosity.
-
I actually compiled it myself using Visual Studio Express Desktop, I'm unsure if there is anything in the compiler that could enable/disable certain things specific to my CPU, but I could get you a link to the binary if you like)
Try this:
http://www.filedropper.com/jhprotominer_1 (http://www.filedropper.com/jhprotominer_1)
(I had an issue trying to get it back down again, but it might be the captcha is case sensitive with no failure error)
All that's been done to it is the VSE 'I need to upgrade this project' (so whatever it did in the background to optimise the code) and built.
Love to know how you go :D -- although you're limited in that if you run 32 threads you won't get 1024MB :P but you can mix and match or just try some high memory ones.
-
@MisO69: Oh and if you've got physical/bios KVM access you might want to see how the performance goes with HT turned off, with this program I expect it's hitting the CPU cores at such a statically high rate there's not much room for HT to actually help (unless you've already done it, then feel free to ignore me)
-
Additional tests:
Found a link to a new 'optimised' (by CPU) miner on YPool based on jhProtominer.
Available here: https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz (https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz)
EDIT: Fixed the link, I gave the wrong one. (There was an extraneous 'p' on the end)
I found my Core i7 went from a levelled 98c/min (after a few hours) to 133c/min with the 'Corei7' optimised version (note: no AVX)
Azure miners didn't see much of a change (though all windows miners are limited to a max of 1024MB of RAM usage instead of Tydus' 4096MB)
-
Additional tests:
Found a link to a new 'optimised' (by CPU) miner on YPool based on jhProtominer.
Available here: https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTzp (https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTzp)
I found my Core i7 went from a levelled 98c/min (after a few hours) to 133c/min with the 'Corei7' optimised version (note: no AVX)
Azure miners didn't see much of a change (though all windows miners are limited to a max of 1024MB of RAM usage instead of Tydus' 4096MB)
It was removed before I got a chance to try it out, do you have another link for it?
It works now, thanks. I'm testing on my cpus now and will post results in a few hours.
-
Additional tests:
Found a link to a new 'optimised' (by CPU) miner on YPool based on jhProtominer.
Available here: https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz (https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz)
EDIT: Fixed the link, I gave the wrong one. (There was an extraneous 'p' on the end)
I found my Core i7 went from a levelled 98c/min (after a few hours) to 133c/min with the 'Corei7' optimised version (note: no AVX)
Azure miners didn't see much of a change (though all windows miners are limited to a max of 1024MB of RAM usage instead of Tydus' 4096MB)
Link disabled. can someone up it to mediafire or something?
-
I've only got the one I downloaded to begin with (Core i7, no AVX) which contains multiple binaries based on memory footprint per thread.
http://www.filedropper.com/jhprotominer-yvg1900-m7c-win64-corei7 (http://www.filedropper.com/jhprotominer-yvg1900-m7c-win64-corei7)
If you're not on a recent (ish) Core-i architecture this probably won't work / be useful for you.
-
I've only got the one I downloaded to begin with (Core i7, no AVX) which contains multiple binaries based on memory footprint per thread.
http://www.filedropper.com/jhprotominer-yvg1900-m7c-win64-corei7 (http://www.filedropper.com/jhprotominer-yvg1900-m7c-win64-corei7)
If you're not on a recent (ish) Core-i architecture this probably won't work / be useful for you.
Awesome, thanks! Getting an additional 50 cpm :)
-
Wow, nice work.
Can you post your before and afters including spec?
-
Only got it up on my laptop so far. Working on getting the VPS's up.
i7 3630QM @ 2.4 GHz:
110 collisions/min : jhProtominer 0.1c -t 8 -m512
163 collisions/min : Win64-corei7-512M -t 8
-
106 cpm i7-870 8 threads -m512
126 cpm i5-3570k stock, 4 threads -m512
55 cpm Phenom 9650 Quad, -m512
111 cpm 2x xeon X7460, 12 threads -m512
118 cpm i5-4430, 4 threads -m512
66 cpm AMD X4 460, 4 threads -m512
73 cpm i3-3220, 4 threads -m512
74 cpm i3-4130, 4 threads -m512
110 cpm FX-8350, 8 threads -m256
127 cpm i7-2600k stock, 8 threads -m512
-
I've only got the one I downloaded to begin with (Core i7, no AVX) which contains multiple binaries based on memory footprint per thread.
http://www.filedropper.com/jhprotominer-yvg1900-m7c-win64-corei7 (http://www.filedropper.com/jhprotominer-yvg1900-m7c-win64-corei7)
If you're not on a recent (ish) Core-i architecture this probably won't work / be useful for you.
Virus total says its clean, still - better run it sandboxed or not on wallet machine...
https://www.virustotal.com/de/file/769840702fa01879feb286295088b0b4292c3373345acb7734fb698155d62225/analysis/1384779805/
-
Yeah, it's up on ypool in the 'how to' so assuming its trusted.
But never know what I've done to it since I downloaded it ;P
-
Any chance anyone has the full download of the AMD and intel versions?
-
Yeah, it's up on ypool in the 'how to' so assuming its trusted.
But never know what I've done to it since I downloaded it ;P
Oh neat, too bad the ypool link doesnt work anymore ^^
-
https://mega.co.nz/#F!h0tkXSxZ!f62uoUXogkxQmP2xO8Ib-g
-
AMD Phenom x4 965 unlocked to a x6 (1645) @ 3.4 Ghz - 130 c/m
-
i7 4770k@4.5GHz, about 270 cpm. -t 8 -m512
I use yvg1900's M7gf Win64 corei7avx
-
For those worrying on authentity of miner software - there are MD5 has sums file signed by private key of developer.
Also M7g for Mac version available.
yvg1900
-
Xeon L5639 @ 3.52ghz does ~280 cpm with -t 11 and 512mb/thread. M7h SSE4 version. Can do more threads but it slows down ltc mining on the same machine.
*12/11/13 updated cpm.
-
Someone with a AMD FX-9590 needs to post their scores. I'm curious what the CPM would be.
-
M7gs-win64-bulldozerv1
FX-8350 @ 5.2Ghz/16Gb 240 7 threads –m512
M7gs-win64-corei7avx
i-7 3770k @ 4.5Ghz/16Gb 300 8 threads –m512
-
In fact, I don't know what edition I used... from this link http://pan.baidu.com/s/11kiAw
Xeon E5-2680v2 @ 3.1GHz 525 40 threads -m1024
-
drmyfore,
You can achieve better results with outher versns I believe.
yvg1900
-
@yvg1900
yeah, I think I can achieve better results, so could you give me some suggestion?
I found there are 2 edition and which edition have 2 version (gs and gf)...
Need I test 18 times to get the fastest miner?
-
"f" vs "s" described in finetuning.txt, for your mach use corei7avx version, 512M is the best I think. Use M7gf in your case.
yvg1900
-
Well, I had thought the bigger memory the better performance...
So what is the max memory per thread of your edition?
-
Using Bulldozer Win64
AMD FX-6200 @ 3.9 GHz/8Gb Averaged ~95 6 threads -m64
AMD FX-6200 @ 3.9 GHz/8Gb 126 6 threads -m128
AMD FX-6200 @ 3.9 GHz/8Gb 132 6 threads -m512
AMD FX-6200 @ 3.9 GHz/8Gb 151 6 threads -m1024
*Please note: Stock clock is 3.8 GHz but results are very similar give or take 2 - 3 CPM average
-
New version available - now matching new pool server configuration after it moved.
Authentic builds are:
https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz
https://mega.co.nz/#F!h0tkXSxZ!f62uoUXogkxQmP2xO8Ib-g
http://sdrv.ms/18wfjYX
http://pan.baidu.com/s/1xOOWa
yvg1900
-
Hi, i dont understand what mins after collision/minut :
i'm a little newby :-)
this is my algo mining:
(http://imageshack.com/a/img12/6394/ul6q.jpg)
thanks!
-
e is standard deviation of CPM value. e of 1.6 is relatively high and is not yet stabilized to make decisions on performance. In your case this is because of finetuning is still in progress and performance varies alot.
AV is algorithm variation. There are 32 significantly different algorithms.
LRT is last round time measured in milliseconds. Another performance metric used to select fastest AV.
After finetuning is complete, you can specify selected AV as a value of -F to skip finetuning for next runs.
yvg1900
-
To add to the list:
- Intel Core 2 Quad Q8300 - 70c/m on average after ~24 hours of use.
- Intel Pentium G860 - 50c/m on average after ~24 hours of use
- Intel Core 2 U9400 - 25c/m on average after ~24 hours of use
-
@yvg1900
I got a haswell i7 4770 which suppot AVX2. And I tried jhprotominer-yvg1900-M7h-win64-coreavx2 and jhprotominer-yvg1900-M7h-win64-corei7avx with the same configurations(8T 512M) but got no difference.
Would you please tell the difference between coreavx2 and corei7avx? And is there any possible to publish a version which support haswell avx2?
thx.
-
coreavx2 IS a version for Haswell.
There are cases with 4770 CPUs when avx/avx2 show no advantage to sse4. In this case your memory subsystem/setup is a bottleneck, not the CPU.
yvg1900
-
thank you and I will try later yvg1900
-
my first try with my 4 Opterons :
- AMD Opteron 6164 HE 70c/m ~24 hours of use
12 thread and 256 MB per thread
That's 280c/m in total. Gonna try different settings and overclock soon.
-
What are "collision per minute" ... I'm use to hashes per second.
Also, does anyone know how to install this on linux and run?
Thanks for any pointers. These new alt coin forums are sometimes a bit scattered.
-
What are "collision per minute" ... I'm use to hashes per second.
Also, does anyone know how to install this on linux and run?
Thanks for any pointers. These new alt coin forums are sometimes a bit scattered.
Taken from this thread: https://bitsharestalk.org/index.php?topic=722.0
sudo apt-get update
sudo apt-get install yasm -y git make g++ build-essential libminiupnpc-dev
sudo apt-get install -y libboost-all-dev libdb++-dev libgmp-dev libssl-dev dos2unix
git clone https://github.com/thbaumbach/ptsminer
cd ~/ptsminer/src
make -f makefile.unix
./ptsminer <your PTS address> <number of threads>
Unsure of what "Linux" you need instructions for, but the above works flawlessly on most Ubuntu 13.04LTS VPS instances.
-
New version available - now matching new pool server configuration after it moved.
Authentic builds are:
https://www.dropbox.com/sh/jvp4wwek8jpohj7/RlW6hzYqTz
https://mega.co.nz/#F!h0tkXSxZ!f62uoUXogkxQmP2xO8Ib-g
http://sdrv.ms/18wfjYX
http://pan.baidu.com/s/1xOOWa
yvg1900
Hello,
Can I have a small feature request? As I'm trying to mine behind a firewall, all my traffic needs to go through a proxy server. It will be really great if jhProtominer supports proxy server. I am willing to donate my first 24hr of mining if I can get it going. :)
-
Will consider proxy support.
yvg1900
-
Will consider proxy support.
yvg1900
Thank you very much for considering my request!! I have a lot of 32 core Xeon servers but they are in a data center behind a firewall. :(
-
AMD FX 8120 - 200 cpm 8 threads (linux)
AMD phenom II x4 - 110 cpm 4 threads (linux)
-
Hi yvg1900,
I'm real newbie and maybe my questions already have been aswered, but I'd like to have your opinion about configuration.
First of all I'm using an Intel® Xeon® E3 1230 v2 (4 coeurs 8 threads @3.3 GHz)
In your finetuning.txt you're talking about the benefits of using separate process per thread using numactl on a Linux but in the same time the new version of jhprotominer (jhprotominer-yvg1900-M7h-linux64-corei7avx.tgz) introduced the "-A" option to enforce cpu affinity.
For you, is it more efficient to run 1 jhprotominer's process using "-A -t 8" or 8 separate process stuck directly to one of the 8 threads?
Thanks in advance if you have the time to respond ands thanks a lot for you're great job!
Pap
-
Two processes make sense only for multi-socket systems (Duals, Quads, SGI UV, etc).
-A option added to lock threads to cores and minimize cross-core thread context migration. This helps in some cases and added to those who like to squeeze every single bit fo perf.
8 separate processes does not make sense at all. I am thinking on better internal NUMA support, too.
The main downside of running multiple processes is that every miner process shall maintain separate pool connection, while threads inside the process share one. Implementing proper NUMA support for single process is a non-trivial task, because of most important point is to get NUMA-aware memory allocation. This is on the list, but needs time for testing.
yvg1900
-
Thank you for you quick response!
So for now is it better to stay with a single process with "-A -t 8" or regarding to your answer maybe 4 process with something like an equivalent to "-A -2" (using numactl or so) to optimize the 1 core/two threads design of the cpu?
Pap
-
And a last question if I can :)
Do you think there's a significant difference between 1024Mo/thread and 2048Mo/thread? And does more Mo/thread would be efficient?
Thank you so much in advance!
Once I'll reach my first PTS (psychological gap for me) I'll double the default time mining for you each hour.
Pap
-
After 512M typically memory bandwidth becomes bottleneck, but you shall try. Advise is very simple - try 512 and 1024, if 1024 is slower then 2048 will be even slower. If 1024 wins - try 2048. Stick with the fastest.
That's it.
yvg1900
-
Anyone been able to connect to http://ptsweb.beeeeer.org pool using yvg1900 miners? Haven't had any success with
jhprotominer.exe -o ptsweb.beeeeer.org:1337 -u PueJycDn6HC2WjzotSs7JjFSYrTGLqSmoG -p x -t 7
-
Thanks for your two replies yvg1900!
I now much better understand the way things goes with your miner.
I've tried by myself and with your precious hints I've optimised not so bad my configuration.
Thank you!
Pap.
-
@ yvg1900
i try your miner and find shares and see the worker at ypool, but in the console it says 0 shares accepted. At ypool no shares/h for this worker. Is it normal or do i have to change something?
Thanks
-
CPU Model/RAM - Collisions per minute - Miner settings
iMac Core i5 680 @ 3.6Ghz/8Gb 61 4 threads –m512
AWS c.3.4large Xeon E5-2680 @2.7Ghz/30Gb 225 4 threads -m512
Azure ExtraLarge AMD 4171 @ 2Ghz/14Gb 110 4 threads -512
Needless to say, Azure are the worst, however with 20 core, you can run 2 ExtraLarge's and a Large. I'm currently getting about 280 c/ms combined
-
Amazon AWS cc2.8xlarge, 32 Threads, 279 cpm jhProtominer -m512
Amazon AWS cc2.8xlarge, 32 Threads, 309 cpm ptsminer 0.5 based on jh's
cc2.8xlarge, dual Xeon E5-2670, 847 cpm jhProtominer-yvg1900 32Threads * 1024MB
Update: updated miner, ~900cpm now.
Update: updated miner again , ~955cpm now.
Update: updated miner again , ~985cpm now.
On my other machine. single Xeon E3-1230 V3, 8Threads * 1024MB. 336 cpm
-
Amazon AWS cc2.8xlarge, 32 Threads, 279 cpm jhProtominer -m512
Amazon AWS cc2.8xlarge, 32 Threads, 309 cpm ptsminer 0.5 based on jh's
cc2.8xlarge, dual Xeon E5-2670, 847 cpm jhProtominer-yvg1900 32Threads * 1024MB
Update: updated miner, ~900cpm now.
Update: updated miner again , ~955cpm now.
How are you getting that much? I'm only getting about 440c/m with the cc2.8xlarge servers. I'm using the core2-1024m miner too..
-
E3-1230V2, win64, -t 8 -m512 -F 0
220cpm
-
E3-1230V2, win64, -t 8 -m512 -F 0
220cpm
which version of jhprotominer are you running?
-
E3-1230V2, win64, -t 8 -m512 -F 0
220cpm
which version of jhprotominer are you running?
the newest one, jhprotominer-yvg1900-M7h-win64-corei7avx,
I just had a look at the output, the fine tuning is completed, and the speed is 241cpm:
collisions/min: 241.5380, e=0.1800; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6493
collisions/min: 241.5375, e=0.1793; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6538
collisions/min: 241.5270, e=0.1785; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6569
collisions/min: 241.5415, e=0.1773; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6582
collisions/min: 241.5610, e=0.1769; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6510
collisions/min: 241.5455, e=0.1756; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6911
collisions/min: 241.5299, e=0.1739; Shares found: 200, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6555
Share found!
collisions/min: 241.5093, e=0.1720; Shares found: 201, submitted 200, accepted 1
97; fine tuning: COMPLETE, AV=29, LRT=6675
collisions/min: 241.5088, e=0.1698; Shares found: 201, submitted 201, accepted 1
98; fine tuning: COMPLETE, AV=29, LRT=6593
-
figured i would try this, but, wow, this miner is horrid
the pool in Austria gets frequent disconnects from 5 different hosts, the miner spends much time in 'warmup' mode, any extra speed is wasted paying to someone else
going through and fixing everything, lol
collisions/min: 224.8938, warmup; Shares found: 12, submitted 12, accepted 12
collisions/min: 233.0360, warmup; Shares found: 21, submitted 21, accepted 21
collisions/min: 192.7183, e=1.2169; Shares found: 24, submitted 24, accepted 21
Will mine 13 more seconds for miner developers to support development of the next version
gg
oh, enjoy my .05 pts that i can never withdraw
it's like i've been violated, the chicanery and swindling that's going on on this pool + miner, scary shit
-
Core i7 2600K @ 4.2Ghz/8Gb 253 8 threads corei7avx-512m -F 0
-
What was the final AV after finetuning swithced to COMPLETED?
yvg1900
-
What was the final AV after finetuning swithced to COMPLETED?
yvg1900
Now - Core i7 2600K @ 4.2Ghz/8Gb 258 8 threads corei7avx-512m -F 0
e=0.0439
AV=5
-
What was the final AV after finetuning swithced to COMPLETED?
yvg1900
sometimes 5 sometimes 29
4670K@3.8G 512M t4 F5 about 250cpm
4770K@3.9G 512M t8 F29 about 278cpm
-
If this is for different t settings, then this is consistent with the concept. If it selects different AVs for same setup, CPM shall be very very close for those AVs.
yvg1900
-
great work!
Can you add the proxy support for jhProtominer?
-
Core i7 2600K @ 4.2Ghz/8Gb 253 8 threads corei7avx-512m -F 0
that's less than I get on 2600K with modified ptsminer
does memory quality make a difference?
-
Dual XEON L5639 @ 2.13G / 16GB 23 threads yvg1900-M7h-linux64-corei7sse4-512M, after fine tuning complete, get 316CPM
AMD FX6200 @ 3.8G / 16GB 5 threads yvg1900-M7h-linux64-bulldozerv1-512M, after fine tuning complete, get around 110CPM
AMD FX8350 @ 4.0G / 16GB 7 threads yvg1900-M7h-linux64-bulldozerv1-512M, after fine tuning complete, get around 130CPM
-
2.4GHz Intel Core 2 Duo (2007 iMac): ~45 c/m (512 mb t -4)
Digital Ocean 2/2 plan: ~30 c/m (256mb t -7)
-
use new version,
Error: boost::bad_any_cast: failed conversion using boost::any_cast
????
-
You have error in params. Post more details.
Sent from my AT10PE-A using Tapatalk
-
C:\Users\Administrator>C:\Windows\coin\c2\yam.exe -M xpt2h://yvg1900.defpts:x@mi
ning.ypool.net:10034:8080:8081:8082:8083:8084:8085:8086:8087/pts
YAM - Yet Another Miner by yvg1900
yam M7i-win64-sandy-bridge/yvg1900
***************************************************
* Supported coins: PTS *
* Author: yvg1900 (Twitter @yvg1900) *
* XPT protocol: jh (http://ypool.net) *
* *
* Addresses for Thanks and Donations: *
* PTS: PZxsEQoiMeB6tHcW2ZySBEiCPio1WkxbEL *
* XPM: AW2388DEWNEfMH4rP9kcj9yKcMq1QywYT4 *
* DTC: D6PmUogMigWvXurgFTqm5VLxQeVpXdYQj3 *
* LTC: Lby4YjhcAxhmbsdHFb4nYydrwGoiJezZt1 *
* BTC: 1FxekeK5La7AuF3oxiLzPKnjXyLMrux6VT *
* NMC: N9KXqmzEqP7gB2dGHpEZiRMgFjUHNM38FR *
***************************************************
Checking PTS optimizations compatibility...
OK: PTS optimizations are compatible
Checking target [xpt2h://yvg1900.defpts:x@mining.ypool.net:10034:8080:8081:8082:
8083:8084:8085:8086:8087/pts] OK
Error: boost::bad_any_cast: failed conversion using boost::any_cast
C:\Users\Administrator>
-
You have error in params. Post more details.
Sent from my AT10PE-A using Tapatalk
hi yvg1900
now I'm trying your new miners but got one question.
I tried yam-yvg1900-M7i-win64-haswell on intel i5 4670K and i7 4770K, there always says
"Error: Can not enable [SeLockMemoryPrivilege] privilege, error=1300
Warning: Huge Pages not available, performance may degrade"
and then I tried it on xeon E3-1245V3 and got no error tips.
so, this is due to diff cpus so i cannot use huge page memory?
(ps.I have enabled the selockmemoryprivilege function in all my Windows 7 x64)
-
Odie2, did you try to run with config?
Sent from my AT10PE-A using Tapatalk
-
Sohueasy, 1300 means that Windows still does not grant you permission. This can be different depending on OS setup. On clean OS install I add Administrators group, current user, disable UAC and then it runs. It is known that enabling huge pages is a mess in Windows.
Sent from my AT10PE-A using Tapatalk
-
hi yvg1900, thx I try to use cfg ,can run :-*
-
Sohueasy, 1300 means that Windows still does not grant you permission. This can be different depending on OS setup. On clean OS install I add Administrators group, current user, disable UAC and then it runs. It is known that enabling huge pages is a mess in Windows.
Sent from my AT10PE-A using Tapatalk
thank you
I disabled UAC and the 1300 error disappeared.
but now it says:
"Warning: ProtoShares memory block for thread 6 allocated in non-hugepages memory
- thread performance may degrade
Error: Can not allocate hugepages memory, error=1450"
does this means I should re-install my win7? or there is no solution to sovle this on win7?
-
Error 1450 means not enough system resources. Try with 1 thread and 512 mob and report how much physical ram do you have. Explain also what you are trying to run - ram/threads wise.
Sent from my AT10PE-A using Tapatalk
-
Error 1450 means not enough system resources. Try with 1 thread and 512 mob and report how much physical ram do you have. Explain also what you are trying to run - ram/threads wise.
Sent from my AT10PE-A using Tapatalk
I reboot my laptop and got results:
when I was trying 512MB/7T before reboot win7, all threads give error 1450
after I reboot win7, only 2 threads give error 1450
it's significantly that windows is a mess...
-
I use your cfg file but get 1300 error.
I am using i5 2500k on Windows 8.1 and UAC is disabled - what should I do to get rid of the error ?
-
Sohueasy - asking once again - how much physical ram your system has? How much is free right after system boot?
Sent from my AT10PE-A using Tapatalk
-
I use your cfg file but get 1300 error.
I am using i5 2500k on Windows 8.1 and UAC is disabled - what should I do to get rid of the error ?
run yam.exe as administator, 1300 error will disappear
-
Krpetrov, there is clear explanation in the docs coming with the miner. You have to grant extra privileges to the user app runs under.
Sent from my AT10PE-A using Tapatalk
-
Thanks man. I solved it.
Any chance for 32bit ver of your miner to pops out ?
-
Sohueasy - asking once again - how much physical ram your system has? How much is free right after system boot?
Sent from my AT10PE-A using Tapatalk
system info: i7 3632qm, 8GB physical mem without any virtual mem
before restart, I got 4.8GB free mem, after reboot I got 5.7GB free mem
yesterday I tried several times,sometimes yam can start with no error, sometimes with 1450 error,it seems all at random
ps. I run 7 threads with 512MB mem each
-
Huge pages assume it is possible to allocate requested memory region in physical ram aligned to specific addresses, and the block shall be continuous. I can easily imagine that even after boot your memory allocation is already somehow fragmented. Also 8gb is absolute minimum, below windows even does not switch huge pages on. Actually, these error messages are just warnings so you take more care of your setup, but only way to really get rid of them is to install more ram...
Follow @yvg1900 on Twitter to get updates on performance mining software
-
Does anybody on here know how many collisions per minute you would need to generate at least one protoshare per day? Also, anyone know how long it would take you to find a block with 20 collisions per minute? Thats what I have on my desktop.
-
I am getting Error: boost::bad_any_cast: failed conversion using boost::any_cast trying to mine protoshares with yam-yvg1900-M7i-win64-sandy-bridge miner any idea whats wrong?
-
@yvg1900
Hi, my cpus are two E5-2680v2
at m7h and use -f 0, I found the best performance is -f 29, and I can catch about 560 at 40T 1024M
TODAY I download the yam edition,
I run two programs at each socket:
1.
PTS Agg. CPM: 611.4; Rnds C/I: 8579/300, Don. C/I: 99/0; Cfg/Thr CPM: 633.3/33.0
66 262 rnds AV=5, ART=7618; FT AV/RT: 5/7601, Best AV/RT: 2/7361
112.124.13.238: On-line, Shares Submitted 517, Accepted 511
Share found!
PTS Agg. CPM: 610.6; Rnds C/I: 8599/300, Don. C/I: 99/0; Cfg/Thr CPM: 609.3/32.2
48 282 rnds AV=5, ART=7627; FT AV/RT: 5/7614, Best AV/RT: 2/7361
112.124.13.238: On-line, Shares Submitted 518, Accepted 512
2.
PTS Agg. CPM: 619.7; Rnds C/I: 9939/320, Don. C/I: 198/0; Cfg/Thr CPM: 589.9/29.
043 277 rnds AV=29, ART=7592; FT AV/RT: 29/7660, Best AV/RT: 11/7242
112.124.13.238: On-line, Shares Submitted 548, Accepted 542
Share found!
Share found!
PTS Agg. CPM: 619.6; Rnds C/I: 9959/320, Don. C/I: 198/0; Cfg/Thr CPM: 587.0/29.
528 297 rnds AV=29, ART=7594; FT AV/RT: 29/7657, Best AV/RT: 11/7242
112.124.13.238: On-line, Shares Submitted 550, Accepted 544
1 + 2>1200?!
FROM 560 TO 1200? So strange!
but It seemed each program can't find the best av value. Please help me, Thanks.
-
by the way,
do you have more than 1024m edition? or does bigger memory for per thread catch the better performance?
because I have 128GB memory, but now I just use 40GB, so...
-
Drmyfore,
Your Mach shall be at approx. 1200-1300 CPM. I test on dual E5-2697v2 and easily go above 1500, but my mach is OCed.
It is also for sure possible that you were significantly underperforming if was running 1 process and no proper numa setup - all the mem accesses were over QPI and extremely slow.
Also this is typical for multi socket machs to fine tune to different av on different socket due to slightly different periferials connected to each socket that lead to different memory controller configs and memory layouts.
You can try to investigate low level hw cfg of your mach and I believe it will explain a lot.
Follow @yvg1900 on Twitter to get updates on performance mining software
-
As of memory edition, Linux versns support up to 2048M, but in your case try 512, compare perf and report. Men bandwidth is a bottleneck, not the size.
Follow @yvg1900 on Twitter to get updates on performance mining software
-
@yvg1900
thanks for you reply.
I use following command to run two programs:
start /high /node 0 /affinity FFFFF run1.bat
start /high /node 1 /affinity FFFFF run2.bat
My memory run at four channels at 1866 per CPU (4DIMMS per CPU)
My MB is ASUS Z9PA-D8
http://www.asus.com/Commercial_Servers_Workstations/Z9PAD8
expect for comparing different memory size, do you think what I need to do adjustments?
-
You shall enable huge pages as explained in the docs, then test 512 and 1024. Run commands seem to be ok, but I would use start directly on miner executable, not on bat file.
Follow @yvg1900 on Twitter to get updates on performance mining software
-
@yvg1900
yeah, I have already enabled huge pages,
I use bat files to ensure that each process can open at the same time
start command can ensure that each process set the corresponding Node
I will test 512, the 1024 seemed about 1250 CPM
-
Server 2003 R2 X64
DUAL E5-4607 /16G jhprotominer 165cpm -t 20 -m512
-
@drmyfore Can you tell me how you set up your system? I have the same hardware(2x 2680v2), but I only get half of your cpm.
I bound 2 processes to each socket/node using numactl, but that didnt help. I'm using the yam-ivy-bridge miner.
-
@EvilTwin
Emm, my operating system is Windows 8.1 Enterprise,
I use "start" to run each bat file.
first, I edit the config files for every process, and create run1.bat/run2.bat
and in order to assign each process to the corresponding Node, I find the "start" command can help me.
for example,
start /high /node 0 /affinity FFFFF run1.bat
start /high /node 1 /affinity FFFFF run2.bat
node parameter is to set the process to corresponding Node.
and the affinity parameter is to set the processor affinity,
you know, the one 2680v2 have 10C20T, so the bin code is 11111111111111111111 (20 Threads, each thread is enabled) , so the in hex it's FFFFF
by the way, if you want to disabled a core, you can change 1 -> 0.
I found you are using Linux, so my solution may not help you. :-[
-
@drmyfore Can you tell me how you set up your system? I have the same hardware(2x 2680v2), but I only get half of your cpm.
I bound 2 processes to each socket/node using numactl, but that didnt help. I'm using the yam-ivy-bridge miner.
You could try just setting affinity in something like htop?
Never done that myself, I'm assuming it works
-
I have a question - can anybody give me a link where I can download the version
jhprotominer-yvg1900-M7h-win64-corei7sse4
or send it to me?
If I try to use the new yam-version my system gives me an error message "the procedure entry point strcpy_s could not be located in the dynamic link library msvcrt.dll". My System is Windows 2003 Server R2
@yvg1900: do you have any news regarding this problem (I'm MoneyMakervB from Twitter).
-
I am preparing new build and this issue (as well as testing on 2003) is on the list.
-
@yvg1900: thanks for the quickly reply. Do you have any suggestions about the time this will need?
-
Win 7 64 bits + yam-yvg1900-M7i-win64-sandy-bridge - i3 2120 - 3.30ghz x 4 = 108 CPM
Ubuntu 13.10 64 bits + yam-yvg1900-M7i-linux64-sandy-bridge - i3 2120 - 3.30ghz x 4 = 138.8 CPM
Ubuntu +29% CPM
-
@yvg1900: thanks for the quickly reply. Do you have any suggestions about the time this will need?
Fix is available. Check your PM, test and post on results.
-
I am just testing the fixed version. Now it is running under Windows 2003 Server R2.
I will let it run overnight and give a statement later on.
@yvg1900: Great job you are doing there
-
I have tried all versions of jhproto miner,but none of them work including the yam version, my system is fedora 20 on IBM POWER 780. Anyone can help me?
-
@yvg1900:
I've tried it now for 9 hours (yam-yvg1900-M7i-win64-nehalem on a Windows 2003 Server R2 with Xeon 5639 processors and 12 GB RAM).
It's running fast (22 threats 512 MB = 386 cpm) but the optimization of av does not stop. It seems that it's trying from AV=1 to AV=32 and then starts once again with AV=1.
For comparison: jhprotominer-yvg1900-M7h-win64-generic has under same configuration (22 threats 512 MB) only 270 cpm.
Greetings
Joerg
-
PM me a log fragment related to fine tuning continuation (I e when it starts over 32 to 1).
Follow @yvg1900 on Twitter to get updates on performance mining software
-
@yvg1900
Plz pm me link yam for server 2003!
Thanks !
-
Can you give me instructions how to create a log? Or only a screenshot? That seems to be unpossible.
-
Screen 1 shows a reset of the procedure without any errormessage:
(http://F:\Protoshares\Fehlerscreens\Screen 1.bmp)
Screen 2 shows a rest of the procedure with a named error.
(http://F:\Protoshares\Fehlerscreens\Screen 2.bmp)
I hope this will help you.
-
Images do not display properly...
Can you provide them alternate way?
-
Note that connection errors that happen during finetuning make finetuning information significantly distorted, so it is by-design behaviour to restart finetuning process if connection error has been detected.
Unfortunately, it is not possible to detect at which stage connection became invalid, if this produced extra CPU load or introduced scheduling pauses.
You can perform finetuning at the pool with high connection stability and use AV parameter with any other pool, because of mining algo is pool neutral.
yvg1900
-
Thanks for the information. I am just testing it.
-
@yvg1900
You are right. With a stable connection the finetuning came to an end and to a result. Now it is running fine. Thanks for the help.
-
Wow, a FX8350 does that well in cpm? Even better than a i7? Or is it the ghz and cores have the biggest impact on cpm?
-
@yvg1900: one question: do you think it is possible to install yam as a service under windows?
-
I think so, I believe there are tools to install any command line app as a service.
Btw, I can not advise from the top of the head which exactly tool shall be used, but I am sure it is possible.
yvg1900
-
excellent
-
@yvg1900:
First of all: thanks a lot for all your efforts!
I'm currently using v1.0e Core2 (512mb) on Centos 6.5 x64 and it's working great! Getting 130 cpm with 2xXeon L5420.
Sadly I can't use the new version of yam. Everytime I try to start it, it tells me "Error: boost::bad_any_cast: failed conversion using boost::any_cast" after "Target OK".
It has to be related to something you changed to the code recently? Anything that can be done to solve the problem?
Thank you again
-
You definitely have some minor mistake/typo in yam config - post it here so we can check it out.
-
Thanks for the answer! :)
My config file looks like this:
threads = 8
mining-params = pts:av=0&m=512
mine = xpt2h://yvg1900.pts_1:x@mining.ypool.net:10034:8080:8081:8082:8083:8084:8085:8086:8087/pts
mine = xpt2://Pm9LE8UxTo5TQZdfH2RqcbNfCBiK1KWCtV:x@112.124.13.238:28988/pts
mine = xram4://Pm9LE8UxTo5TQZdfH2RqcbNfCBiK1KWCtV@ptsmine.beeeeer.org:1337/pts
#proxy = socks4a://127.0.0.1:9150
compact-stats = 1
print-timestamps = 0
(yes, I kept your URIs, just to try if it is working)
when I ran it with: ./yam --config yam-pts.cfg
It tells me: Error: At least one mining target shall be defined with --mine parameter
(shouldn't it take the first parameter from the config?
since this doesn't work, I tried: ./yam --config yam-pts.cfg --mine xpt2h://yvg1900.pts_1:x@mining.ypool.net:10034:8080:8081:8082:8083:8084:8085:8086:8087/pts
and it tells me:
Miner version: yam M7m-linux64-core2/yvg1900
Checking target [xram4://Pm9LE8UxTo5TQZdfH2RqcbNfCBiK1KWCtV@ptsmine.beeeeer.org:1337/pts]...
Target OK
Error: boost::bad_any_cast: failed conversion using boost::any_cast
Maybe I am doing somehting very basic wrong, but I'm just not able to get it to work.
Thank you again for the help!
-
Sooo, I finally got it solved.
Very silly mistake.
./yam --config yam-pts.cfg
you have to include the folder to yam-pts.cfg, for example:
./yam --config yam-yvg1900/linux64-core2/yam-pts.cfg
A "couldn't find the config file" mistake would be very nice for the next version. I hope I could help somebody out.
Anyway, thanks again for your efforts yvg1900, I'm nearly getting 100% improvement with this new version! +5%
-
Noob question: I'm mining with an i7 4770k. When I set the number of cores to 4 I get roughly 150 collisions/m. When I set the cores to 8 I get about 130. Is that the number per core, or the total number of collisions? Trying to figure out which way is more efficient. Thanks.
-
Sooo, I finally got it solved.
Very silly mistake.
./yam --config yam-pts.cfg
you have to include the folder to yam-pts.cfg, for example:
./yam --config yam-yvg1900/linux64-core2/yam-pts.cfg
A "couldn't find the config file" mistake would be very nice for the next version. I hope I could help somebody out.
Anyway, thanks again for your efforts yvg1900, I'm nearly getting 100% improvement with this new version! +5%
Will consider better error messages for M7o.
-
can this miner be used on Freebsd?
-
@yvg1900
thank you for ur effort first!
After i upgrade from M7t to M7v,i found it seems like the settings things have bit difference because i cant use it for mining at all.Could u please teach me how to config it and make it work?thanks a lot.
-
Missed this thread, probably it is not too late...
Is there still a chance to get output of the miner?
And feel free to PM me if you need urgent answer.