Author Topic: [ANN] MMC.1GH.COM - GPU mining, CPU mining ★  (Read 161152 times)

0 Members and 1 Guest are viewing this topic.

Offline reorder

  • Sr. Member
  • ****
  • Posts: 400
    • View Profile
Hi I have two Gigabyte AMD 7970 GPU and the issue I am having is that it will mine for a few hours and then activity for the GPU's on the AMD VISION Engine Control Center goes down to 0%. When it is mining activity is 99%. Then I look over at the command prompt to see if it shows any sort of error and it doesn't show any sort of error

It just shows what the last share was like the hpm and at what time.

Is anyone having this issue or have any idea what is going on? Also clock settings for GPUs are 975 and 1255 for memory.
I'd suspect card hanging due to overheat here.

Offline jae208

  • Hero Member
  • *****
  • Posts: 525
    • View Profile
Hi I have two Gigabyte AMD 7970 GPU and the issue I am having is that it will mine for a few hours and then activity for the GPU's on the AMD VISION Engine Control Center goes down to 0%. When it is mining activity is 99%. Then I look over at the command prompt to see if it shows any sort of error and it doesn't show any sort of error

It just shows what the last share was like the hpm and at what time.

Is anyone having this issue or have any idea what is going on? Also clock settings for GPUs are 975 and 1255 for memory.
http://bitsharestutorials.com A work in progress
Subscribe to the Youtube Channel
https://www.youtube.com/user/BitsharesTutorials

Offline reorder

  • Sr. Member
  • ****
  • Posts: 400
    • View Profile
Guys,

it is now confirmed and official: if you use Windows and get this error:
Code: [Select]
MemoryError: clEnqueueNDRangeKernel failed: mem object allocation failure
your system is lacking free RAM to allocate required amount of VRAM on the card. Your options are either adding more system RAM or switching to Ubuntu.

Thanks to everyone who has helped nailing down this one.

Offline microxp

  • Newbie
  • *
  • Posts: 15
    • View Profile
HIS R9 290X got 20HPM..
the same as 280x..
maybe something is wrong?
290x twice the price of 280x!

Offline reorder

  • Sr. Member
  • ****
  • Posts: 400
    • View Profile
Getting the same error of memory allocation, but with 3GB of a R9 280x


Quote
Skipping device 1 (        Intel(R) Core(TM) i3-2100 CPU @ 3.10GHz
  ): not a GPU
Thu Jan 02 02:52:39 2014 Miner Starts - mmcpool.1gh.com:8080
MC2Hasher:0 initialized on device: Tahiti
                   type: 4, vendor: Advanced Micro Devices, Inc.   , mem: 322122
5472 with 1-pass algo
Exception in thread Thread-1:
Traceback (most recent call last):
  File "threading.pyo", line 808, in __bootstrap_inner
  File "threading.pyo", line 761, in run
  File "miner.py", line 2804, in miner.miner_thread (miner.c:8363)
  File "miner.py", line 2799, in miner.Miner.loop (miner.c:8231)
  File "miner.py", line 2788, in miner.Miner.iterate (miner.c:8056)
  File "miner.py", line 2580, in miner.MC2Hasher.processMidstate (miner.c:4068)
  File "miner.py", line 2570, in miner.MC2Hasher.search (miner.c:3833)
MemoryError: clEnqueueNDRangeKernel failed: mem object allocation failure

Any ideas?
BTW my mobo chipset doesn't have integrated graphics.
How much RAM do you have in the system? Some people in the thread have reported Windows needs >2G system RAM to allocated required videoram.

Offline metraque

  • Newbie
  • *
  • Posts: 14
    • View Profile
Getting the same error of memory allocation, but with 3GB of a R9 280x


Quote
Skipping device 1 (        Intel(R) Core(TM) i3-2100 CPU @ 3.10GHz
  ): not a GPU
Thu Jan 02 02:52:39 2014 Miner Starts - mmcpool.1gh.com:8080
MC2Hasher:0 initialized on device: Tahiti
                   type: 4, vendor: Advanced Micro Devices, Inc.   , mem: 322122
5472 with 1-pass algo
Exception in thread Thread-1:
Traceback (most recent call last):
  File "threading.pyo", line 808, in __bootstrap_inner
  File "threading.pyo", line 761, in run
  File "miner.py", line 2804, in miner.miner_thread (miner.c:8363)
  File "miner.py", line 2799, in miner.Miner.loop (miner.c:8231)
  File "miner.py", line 2788, in miner.Miner.iterate (miner.c:8056)
  File "miner.py", line 2580, in miner.MC2Hasher.processMidstate (miner.c:4068)
  File "miner.py", line 2570, in miner.MC2Hasher.search (miner.c:3833)
MemoryError: clEnqueueNDRangeKernel failed: mem object allocation failure

Any ideas?
BTW my mobo chipset doesn't have integrated graphics.
« Last Edit: January 02, 2014, 05:58:29 am by metraque »

Offline Tartempion

  • Newbie
  • *
  • Posts: 4
    • View Profile
Check the temperature with:

Code: [Select]
aticonfig --adapter=0 --od-gettemperature
and fan speed with

Code: [Select]
aticonfig --pplib-cmd "get fanspeed 0"
If the temperature is too high, try setting the fan speed like this:

Code: [Select]
aticonfig --pplib-cmd "set fanspeed 0 80"
In Windows world they have all those fancy GUIs ;)
my problem is I get
PPLIB command execution has failed!
ati_pplib_cmd: execute "get" failed!

and with cgminer i have no problem with these cards runing at 70% fan speed with 76 degree
thanks for all your help!!

Since aticonfig doesnt allow me to control the fan at the 290 cards, I use now following solution to protect against the temperature run to 94 degree.
I first start cgminer and when card reaches arround 76 degree and 60-70% fan rotation I quickly switch to memory coin miner and the fan stays at that rotation speed.
It seems to me the auto speed of the cards is set way too low around 43 % fan rotation.

Another option is get the source of AMDOverDriveCtrl and patch is as follow:

diff --git a/src/adl.cpp b/src/adl.cpp                                                                 │
index ffb1dea..b9631a1 100644                                                                          │
--- a/src/adl.cpp                                                                                      │
+++ b/src/adl.cpp                                                                                      │
@@ -740,9 +740,9 @@ int ADL::UpdateData()                                                              │
        mODParameters.iSize = sizeof(ADLODParameters);                                                 │
        if (SAVE_CALL(ADL_Overdrive5_ODParameters_Get)(mGPUIndex, &mODParameters) == ADL_OK)           │
        {                                                                                              │
-           if (mODParameters.sEngineClock.iMin == mODParameters.sEngineClock.iMax ||                  │
+           if (/*mODParameters.sEngineClock.iMin == mODParameters.sEngineClock.iMax ||                │
                mODParameters.sMemoryClock.iMin == mODParameters.sMemoryClock.iMax ||                  │
-               mODParameters.sVddc.iMin == mODParameters.sVddc.iMax ||                                │
+               mODParameters.sVddc.iMin == mODParameters.sVddc.iMax || */                             │
                mODParameters.sEngineClock.iMin > mODParameters.sEngineClock.iMax ||                   │
                mODParameters.sMemoryClock.iMin > mODParameters.sMemoryClock.iMax ||                   │
                mODParameters.sVddc.iMin > mODParameters.sVddc.iMax)                                   │

Offline Tartempion

  • Newbie
  • *
  • Posts: 4
    • View Profile
could you add an option to control the vents rotation on card

thanks in advance
Perhaps, in future versions. Driver's autotune might work just as well in most cases: card's cores are not always loaded to 100%, during periods when most time is spent accessing memory, cores have a chance to cool down.

thanks a lot I fear only that 94 degree ist too hot for my 290
Some people have reported a much cooler operation than with scrypt, and a lower power consumption.
yes you are right on my saphire its 67 but on ASUS card its 94 which is very sad
Figure you have to check its fans and thermal grease.. 94 is not healthy at all.

On my R9 290 (Sapphire) in Linux, I observe the same thing. With Catalyst let alone managing the card, it settles around 94C
but the fan is running very slowly, around 50%, maybe less.

I manually ramp them up with AMDOverdriveCtrl (you need a patch to make it work) but the aticonfig command line
should work too.

Note that I'm surprised some report a lower temp on Sapphire, I've checked I'm using the latest Sapphire BIOS
015.041.000.000.000000 which is supposed to improve that...

Offline DWz

  • Newbie
  • *
  • Posts: 3
    • View Profile
Well, I can confirm it's not working on Cayman arch's very well.  My 6970 is getting nothing but rejects, whilst my 7950 TF3's are speeding along beside it on the same board.   

Anyone know offhand what the command prompt syntax in Windoze is for disabling that gpu?  The readme isn't quite explicit enough in that regard. Thanks ;)
Just enumerate device numbers you want to mine on (X in Worker:X string) after your address on the commandline, space-separated.


Copy that, many thanks reorder. :)  When you get the Cayman XT's working proper, the next few rounds of beer's are on me. :D

Offline Trokhon

  • Newbie
  • *
  • Posts: 10
    • View Profile
Check the temperature with:

Code: [Select]
aticonfig --adapter=0 --od-gettemperature
and fan speed with

Code: [Select]
aticonfig --pplib-cmd "get fanspeed 0"
If the temperature is too high, try setting the fan speed like this:

Code: [Select]
aticonfig --pplib-cmd "set fanspeed 0 80"
In Windows world they have all those fancy GUIs ;)
my problem is I get
PPLIB command execution has failed!
ati_pplib_cmd: execute "get" failed!

and with cgminer i have no problem with these cards runing at 70% fan speed with 76 degree
thanks for all your help!!

Since aticonfig doesnt allow me to control the fan at the 290 cards, I use now following solution to protect against the temperature run to 94 degree.
I first start cgminer and when card reaches arround 76 degree and 60-70% fan rotation I quickly switch to memory coin miner and the fan stays at that rotation speed.
It seems to me the auto speed of the cards is set way too low around 43 % fan rotation.

Offline reorder

  • Sr. Member
  • ****
  • Posts: 400
    • View Profile
i redownloaded and got the same error as above.
my command to run miner is :
GPU_MAX_ALLOC_PERCENT=100 GPU_MAX_HEAP_SIZE=100 ./mmc-miner MNafBQWjxf6t2moSuRbpUkG7Uo8m8twWwX
I have another update, would you please try again? (I am trying options that are likely to help blindly as I do not have hardware to test myself, my apologies if it bothers you).

Same error again, i have 7670m with 2GB RAM.

Code: [Select]
~/Desktop/mmc/mc2-1gh-lin-x64$ ./mmc-miner MNafBQWjxf6t2moSuRbpUkG7Uo8m8twWwX
Skipping device 1 (Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz): not a GPU
Wed Jan  1 20:26:56 2014 Miner Starts - mmcpool.1gh.com:8080
MC2Hasher:0 initialized on device: Turks type: 4, vendor: Advanced Micro Devices, Inc., mem: 2147483648 with 2-pass algo
Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib/python2.7/threading.py", line 551, in __bootstrap_inner
    flag that can be set to true with the set() method and reset to false
  File "/usr/lib/python2.7/threading.py", line 504, in run
    given, value defaults to 1.
  File "miner.py", line 2804, in miner.miner_thread (miner.c:7679)
  File "miner.py", line 2799, in miner.Miner.loop (miner.c:7558)
  File "miner.py", line 2788, in miner.Miner.iterate (miner.c:7392)
  File "miner.py", line 2580, in miner.MC2Hasher.processMidstate (miner.c:3524)
  File "miner.py", line 2570, in miner.MC2Hasher.search (miner.c:3297)
MemoryError: clEnqueueNDRangeKernel failed: mem object allocation failure
Basically it means the system cannot allocate 1.5G VRAM in 1G and 0.5G contiguous pieces for miner. Any chance to lower the system VRAM usage? By switching to integrated video for exampe, if your system can do that.

Offline reorder

  • Sr. Member
  • ****
  • Posts: 400
    • View Profile
Hello,

I-m trying to run mmc-miner in Litecoin BAMT rig, I get bash mmc-miner cannot execute binary file. Any help is truly appreciated.

Thanks!
BAMT is 32bit only I believe. Miner binary is 64bit.

EDIT: I have 32bit binary tailored for BAMT in plans, but not sure what to do first - try to iron out Turks issues or that.

Offline metraque

  • Newbie
  • *
  • Posts: 14
    • View Profile
Hello,

I-m trying to run mmc-miner in Litecoin BAMT rig, I get bash mmc-miner cannot execute binary file. Any help is truly appreciated.

Thanks!

Offline kotofey

  • Newbie
  • *
  • Posts: 6
    • View Profile
Is there someone who uses this miner with 4 cards? If yes, which is your config/OS/drivers?

Offline infrablack

  • Newbie
  • *
  • Posts: 4
    • View Profile
i redownloaded and got the same error as above.
my command to run miner is :
GPU_MAX_ALLOC_PERCENT=100 GPU_MAX_HEAP_SIZE=100 ./mmc-miner MNafBQWjxf6t2moSuRbpUkG7Uo8m8twWwX
I have another update, would you please try again? (I am trying options that are likely to help blindly as I do not have hardware to test myself, my apologies if it bothers you).

Same error again, i have 7670m with 2GB RAM.

Code: [Select]
~/Desktop/mmc/mc2-1gh-lin-x64$ ./mmc-miner MNafBQWjxf6t2moSuRbpUkG7Uo8m8twWwX
Skipping device 1 (Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz): not a GPU
Wed Jan  1 20:26:56 2014 Miner Starts - mmcpool.1gh.com:8080
MC2Hasher:0 initialized on device: Turks type: 4, vendor: Advanced Micro Devices, Inc., mem: 2147483648 with 2-pass algo
Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib/python2.7/threading.py", line 551, in __bootstrap_inner
    flag that can be set to true with the set() method and reset to false
  File "/usr/lib/python2.7/threading.py", line 504, in run
    given, value defaults to 1.
  File "miner.py", line 2804, in miner.miner_thread (miner.c:7679)
  File "miner.py", line 2799, in miner.Miner.loop (miner.c:7558)
  File "miner.py", line 2788, in miner.Miner.iterate (miner.c:7392)
  File "miner.py", line 2580, in miner.MC2Hasher.processMidstate (miner.c:3524)
  File "miner.py", line 2570, in miner.MC2Hasher.search (miner.c:3297)
MemoryError: clEnqueueNDRangeKernel failed: mem object allocation failure