Optimize your GPU. Find the value the easy way.


log in

Advanced search

Message boards : Number crunching : Optimize your GPU. Find the value the easy way.

Previous · 1 · 2 · 3 · 4 · 5 . . . 13 · Next
Author Message
Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1275682 - Posted: 26 Aug 2012, 9:02:40 UTC - in response to Message 1275613.


So ... what am I not doing correctly? (Yes, I stopped BOINC Manager before executing the performance utility).

V 1.1 will have a lot more diagnostics.

The files are copied
C:\Users\username\AppData\Roaming\eFMer\SetiPerformance\slotx

In that location the exe is started. Is it there? (You have to press start first to copy the files.)
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1275716 - Posted: 26 Aug 2012, 11:31:38 UTC - in response to Message 1275565.
Last modified: 26 Aug 2012, 11:32:15 UTC


So I'm going to try 4 wu's each on my GPU's (no HT, 2 core's)

0.5 should be the best.


Hi

I´m a x41z user, your build works ok in some hosts but when i try to use in this host: 6690764 (with just one 590 EVGA) its give me 100 sec in all measures (1, 2 or 3 WU), and shows just 1 device (the 2 GPUs on the 590 works OK with Boinc processing 2x2WU at a time XP32 and no OC). Do you know anythink that could cause this?

You can take a look in: C:\Users\username\AppData\Roaming\eFMer\SetiPerformance

The applications are in slotx, but are removed after you press cancel.
In the slot, check the application log for an error
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

juan BFBProject donor
Volunteer tester
Avatar
Send message
Joined: 16 Mar 07
Posts: 5487
Credit: 316,277,715
RAC: 139,512
Brazil
Message 1275727 - Posted: 26 Aug 2012, 12:27:07 UTC - in response to Message 1275716.

Ok, i´m on XP so the path is little diferent but i foud.

WHen runs this is the oputput:


Starting automatic test: (lunatics_x41z_win32_cuda42.exe)
26 August 2012 - 09:17:24 Start, devices: 1, device count: 1
26 August 2012 - 09:19:06 Runtime: Device: 0, count: 0, 101 seconds
26 August 2012 - 09:19:06 Device: 0, finished
Ready ---------------------------------------------------------------------
Results:
Device: 0, device count: 1, average time / count: 101, average time on device: 101 Seconds (1 Minutes, 41 Seconds)
Next ---------------------------------------------------------------------
26 August 2012 - 09:19:06 Start, devices: 1, device count: 2
26 August 2012 - 09:22:31 Runtime: Device: 0, count: 1, 202 seconds
26 August 2012 - 09:22:31 Runtime: Device: 0, count: 0, 202 seconds
26 August 2012 - 09:22:31 Device: 0, finished
26 August 2012 - 09:22:31 Device: 0, finished
Ready ---------------------------------------------------------------------
Results:
Device: 0, device count: 2, average time / count: 202, average time on device: 101 Seconds (1 Minutes, 41 Seconds)
The best average time found: 101 Seconds (1 Minutes, 41 Seconds), with count: 1.00 (1)


Then on slot 0


an't open init data file - running in standalone mode
Can't open init data file - running in standalone mode
setiathome_CUDA: Found 2 CUDA device(s):
Device 1: GeForce GTX 590, 1535 MiB, regsPerBlock 32768
computeCap 2.0, multiProcs 16
pciBusID = 3, pciSlotID = 0
clockRate = 1260 MHz
Device 2: GeForce GTX 590, 1535 MiB, regsPerBlock 32768
computeCap 2.0, multiProcs 16
pciBusID = 4, pciSlotID = 0
clockRate = 1260 MHz
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 590 is okay
SETI@home using CUDA accelerated device GeForce GTX 590
pulsefind: blocks per SM 4 (Fermi or newer default)
pulsefind: periods per launch 100 (default)
Priority of process set to BELOW_NORMAL (default) successfully
Priority of worker thread set successfully

setiathome enhanced x41z, Cuda 4.20

Legacy setiathome_enhanced V6 mode.
Work Unit Info:
...............
WU true angle range is : 0.775000
VRAM: cudaMalloc((void**) &dev_cx_DataArray, 1048576x 8bytes = 8388608bytes, offs256=0, rtotal= 8388608bytes
VRAM: cudaMalloc((void**) &dev_cx_ChirpDataArray, 1179648x 8bytes = 9437184bytes, offs256=0, rtotal= 17825792bytes
VRAM: cudaMalloc((void**) &dev_flag, 1x 8bytes = 8bytes, offs256=0, rtotal= 17825800bytes
VRAM: cudaMalloc((void**) &dev_WorkData, 1179648x 8bytes = 9437184bytes, offs256=0, rtotal= 27262984bytes
VRAM: cudaMalloc((void**) &dev_PowerSpectrum, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 31457288bytes
VRAM: cudaMalloc((void**) &dev_t_PowerSpectrum, 1048584x 4bytes = 1048608bytes, offs256=0, rtotal= 32505896bytes
VRAM: cudaMalloc((void**) &dev_GaussFitResults, 1048576x 16bytes = 16777216bytes, offs256=0, rtotal= 49283112bytes
VRAM: cudaMalloc((void**) &dev_PoT, 1572864x 4bytes = 6291456bytes, offs256=0, rtotal= 55574568bytes
VRAM: cudaMalloc((void**) &dev_PoTPrefixSum, 1572864x 4bytes = 6291456bytes, offs256=0, rtotal= 61866024bytes
VRAM: cudaMalloc((void**) &dev_NormMaxPower, 16384x 4bytes = 65536bytes, offs256=0, rtotal= 61931560bytes
VRAM: cudaMalloc((void**) &dev_flagged, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 66125864bytes
VRAM: cudaMalloc((void**) &dev_outputposition, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 70320168bytes
VRAM: cudaMalloc((void**) &dev_PowerSpectrumSumMax, 262144x 12bytes = 3145728bytes, offs256=0, rtotal= 73465896bytes
VRAM: cudaMallocArray( &dev_gauss_dof_lcgf_cache, 1x 8192bytes = 8192bytes, offs256=184, rtotal= 73474088bytes
VRAM: cudaMallocArray( &dev_null_dof_lcgf_cache, 1x 8192bytes = 8192bytes, offs256=0, rtotal= 73482280bytes
VRAM: cudaMalloc((void**) &dev_find_pulse_flag, 1x 8bytes = 8bytes, offs256=0, rtotal= 73482288bytes
VRAM: cudaMalloc((void**) &dev_t_funct_cache, 1966081x 4bytes = 7864324bytes, offs256=0, rtotal= 81346612bytes
Thread call stack limit is: 1k
cudaAcc_free() called...
cudaAcc_free() running...
cudaAcc_free() PulseFind freed...
cudaAcc_free() Gaussfit freed...
cudaAcc_free() AutoCorrelation freed...
cudaAcc_free() DONE.

Flopcounter: 12287736759182.770000

Spike count: 13
Pulse count: 1
Triplet count: 0
Gaussian count: 0
Worker preemptively acknowledging a normal exit.->
called boinc_finish
boinc_exit(): requesting safe worker shutdown ->
boinc_exit(): received safe worker shutdown acknowledge ->


on slot1


Can't open init data file - running in standalone mode
Can't open init data file - running in standalone mode
Can't open init data file - running in standalone mode
setiathome_CUDA: Found 2 CUDA device(s):
Device 1: GeForce GTX 590, 1535 MiB, regsPerBlock 32768
computeCap 2.0, multiProcs 16
pciBusID = 3, pciSlotID = 0
clockRate = 1260 MHz
Device 2: GeForce GTX 590, 1535 MiB, regsPerBlock 32768
computeCap 2.0, multiProcs 16
pciBusID = 4, pciSlotID = 0
clockRate = 1260 MHz
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 590 is okay
SETI@home using CUDA accelerated device GeForce GTX 590
pulsefind: blocks per SM 4 (Fermi or newer default)
pulsefind: periods per launch 100 (default)
Priority of process set to BELOW_NORMAL (default) successfully
Priority of worker thread set successfully

setiathome enhanced x41z, Cuda 4.20

Legacy setiathome_enhanced V6 mode.
Work Unit Info:
...............
WU true angle range is : 0.775000
VRAM: cudaMalloc((void**) &dev_cx_DataArray, 1048576x 8bytes = 8388608bytes, offs256=0, rtotal= 8388608bytes
VRAM: cudaMalloc((void**) &dev_cx_ChirpDataArray, 1179648x 8bytes = 9437184bytes, offs256=0, rtotal= 17825792bytes
VRAM: cudaMalloc((void**) &dev_flag, 1x 8bytes = 8bytes, offs256=0, rtotal= 17825800bytes
VRAM: cudaMalloc((void**) &dev_WorkData, 1179648x 8bytes = 9437184bytes, offs256=0, rtotal= 27262984bytes
VRAM: cudaMalloc((void**) &dev_PowerSpectrum, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 31457288bytes
VRAM: cudaMalloc((void**) &dev_t_PowerSpectrum, 1048584x 4bytes = 1048608bytes, offs256=0, rtotal= 32505896bytes
VRAM: cudaMalloc((void**) &dev_GaussFitResults, 1048576x 16bytes = 16777216bytes, offs256=0, rtotal= 49283112bytes
VRAM: cudaMalloc((void**) &dev_PoT, 1572864x 4bytes = 6291456bytes, offs256=0, rtotal= 55574568bytes
VRAM: cudaMalloc((void**) &dev_PoTPrefixSum, 1572864x 4bytes = 6291456bytes, offs256=0, rtotal= 61866024bytes
VRAM: cudaMalloc((void**) &dev_NormMaxPower, 16384x 4bytes = 65536bytes, offs256=0, rtotal= 61931560bytes
VRAM: cudaMalloc((void**) &dev_flagged, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 66125864bytes
VRAM: cudaMalloc((void**) &dev_outputposition, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 70320168bytes
VRAM: cudaMalloc((void**) &dev_PowerSpectrumSumMax, 262144x 12bytes = 3145728bytes, offs256=0, rtotal= 73465896bytes
VRAM: cudaMallocArray( &dev_gauss_dof_lcgf_cache, 1x 8192bytes = 8192bytes, offs256=208, rtotal= 73474088bytes
VRAM: cudaMallocArray( &dev_null_dof_lcgf_cache, 1x 8192bytes = 8192bytes, offs256=56, rtotal= 73482280bytes
VRAM: cudaMalloc((void**) &dev_find_pulse_flag, 1x 8bytes = 8bytes, offs256=0, rtotal= 73482288bytes
VRAM: cudaMalloc((void**) &dev_t_funct_cache, 1966081x 4bytes = 7864324bytes, offs256=0, rtotal= 81346612bytes
Thread call stack limit is: 1k
cudaAcc_free() called...
cudaAcc_free() running...
cudaAcc_free() PulseFind freed...
cudaAcc_free() Gaussfit freed...
cudaAcc_free() AutoCorrelation freed...
cudaAcc_free() DONE.

Flopcounter: 12287736759182.770000

Spike count: 13
Pulse count: 1
Triplet count: 0
Gaussian count: 0
Worker preemptively acknowledging a normal exit.->
called boinc_finish
boinc_exit(): requesting safe worker shutdown ->
boinc_exit(): received safe worker shutdown acknowledge ->

setiathome_CUDA: Found 2 CUDA device(s):
Device 1: GeForce GTX 590, 1535 MiB, regsPerBlock 32768
computeCap 2.0, multiProcs 16
pciBusID = 3, pciSlotID = 0
clockRate = 1260 MHz
Device 2: GeForce GTX 590, 1535 MiB, regsPerBlock 32768
computeCap 2.0, multiProcs 16
pciBusID = 4, pciSlotID = 0
clockRate = 1260 MHz
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 590 is okay
SETI@home using CUDA accelerated device GeForce GTX 590
pulsefind: blocks per SM 4 (Fermi or newer default)
pulsefind: periods per launch 100 (default)
Priority of process set to BELOW_NORMAL (default) successfully
Priority of worker thread set successfully

setiathome enhanced x41z, Cuda 4.20

Legacy setiathome_enhanced V6 mode.
Work Unit Info:
...............
WU true angle range is : 0.775000
VRAM: cudaMalloc((void**) &dev_cx_DataArray, 1048576x 8bytes = 8388608bytes, offs256=0, rtotal= 8388608bytes
VRAM: cudaMalloc((void**) &dev_cx_ChirpDataArray, 1179648x 8bytes = 9437184bytes, offs256=0, rtotal= 17825792bytes
VRAM: cudaMalloc((void**) &dev_flag, 1x 8bytes = 8bytes, offs256=0, rtotal= 17825800bytes
VRAM: cudaMalloc((void**) &dev_WorkData, 1179648x 8bytes = 9437184bytes, offs256=0, rtotal= 27262984bytes
VRAM: cudaMalloc((void**) &dev_PowerSpectrum, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 31457288bytes
VRAM: cudaMalloc((void**) &dev_t_PowerSpectrum, 1048584x 4bytes = 1048608bytes, offs256=0, rtotal= 32505896bytes
VRAM: cudaMalloc((void**) &dev_GaussFitResults, 1048576x 16bytes = 16777216bytes, offs256=0, rtotal= 49283112bytes
VRAM: cudaMalloc((void**) &dev_PoT, 1572864x 4bytes = 6291456bytes, offs256=0, rtotal= 55574568bytes
VRAM: cudaMalloc((void**) &dev_PoTPrefixSum, 1572864x 4bytes = 6291456bytes, offs256=0, rtotal= 61866024bytes
VRAM: cudaMalloc((void**) &dev_NormMaxPower, 16384x 4bytes = 65536bytes, offs256=0, rtotal= 61931560bytes
VRAM: cudaMalloc((void**) &dev_flagged, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 66125864bytes
VRAM: cudaMalloc((void**) &dev_outputposition, 1048576x 4bytes = 4194304bytes, offs256=0, rtotal= 70320168bytes
VRAM: cudaMalloc((void**) &dev_PowerSpectrumSumMax, 262144x 12bytes = 3145728bytes, offs256=0, rtotal= 73465896bytes
VRAM: cudaMallocArray( &dev_gauss_dof_lcgf_cache, 1x 8192bytes = 8192bytes, offs256=208, rtotal= 73474088bytes
VRAM: cudaMallocArray( &dev_null_dof_lcgf_cache, 1x 8192bytes = 8192bytes, offs256=56, rtotal= 73482280bytes
VRAM: cudaMalloc((void**) &dev_find_pulse_flag, 1x 8bytes = 8bytes, offs256=0, rtotal= 73482288bytes
VRAM: cudaMalloc((void**) &dev_t_funct_cache, 1966081x 4bytes = 7864324bytes, offs256=0, rtotal= 81346612bytes
Thread call stack limit is: 1k
cudaAcc_free() called...
cudaAcc_free() running...
cudaAcc_free() PulseFind freed...
cudaAcc_free() Gaussfit freed...
cudaAcc_free() AutoCorrelation freed...
cudaAcc_free() DONE.

Flopcounter: 12287736759182.770000

Spike count: 13
Pulse count: 1
Triplet count: 0
Gaussian count: 0
Worker preemptively acknowledging a normal exit.->
called boinc_finish
boinc_exit(): requesting safe worker shutdown ->
boinc_exit(): received safe worker shutdown acknowledge ->


no work was done for the 2 GPU or in the others slots.

maybe?: Can't open init data file - running in standalone mode

Any clue?

Sorry to take your time.
____________

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1275742 - Posted: 26 Aug 2012, 12:56:03 UTC - in response to Message 1275727.


Any clue?

Set devices to 2 and try again.
It's one board but with 2 devices.

For comparison I like to see the automatic test log.
I have a 590 on a Win 7 system.
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

Profile Lint trap
Send message
Joined: 30 May 03
Posts: 871
Credit: 28,061,734
RAC: 10,444
United States
Message 1275767 - Posted: 26 Aug 2012, 14:28:14 UTC


GTX 460 o/c'd, 1gb, 256-bit. XPP-sp3 32-bit, NV driver 301.42, system idle.


Starting test: (lunatics_x41g_win32_cuda32_perf.exe)

Device: 0, device count: 1, average time / count: 162, average time on device: 162 Seconds (2 Minutes, 42 Seconds)
Device: 0, device count: 2, average time / count: 317, average time on device: 158 Seconds (2 Minutes, 38 Seconds)
Device: 0, device count: 3, average time / count: 474, average time on device: 158 Seconds (2 Minutes, 38 Seconds)
Device: 0, device count: 4, average time / count: 635, average time on device: 158 Seconds (2 Minutes, 38 Seconds)


I was surprised to see the 460/this system 'appears' to have more headroom available after posting identical averages for 2, 3, and 4 counts...



Lt



juan BFBProject donor
Volunteer tester
Avatar
Send message
Joined: 16 Mar 07
Posts: 5487
Credit: 316,277,715
RAC: 139,512
Brazil
Message 1275831 - Posted: 26 Aug 2012, 16:37:43 UTC - in response to Message 1275742.
Last modified: 26 Aug 2012, 16:40:16 UTC


Any clue?

Set devices to 2 and try again.
It's one board but with 2 devices.

For comparison I like to see the automatic test log.
I have a 590 on a Win 7 system.


Hi this is what you need?

Starting automatic test: (lunatics_x41z_win32_cuda42.exe)
26 August 2012 - 13:28:26 Start, devices: 2, device count: 1
26 August 2012 - 13:30:23 Runtime: Device: 0, count: 0, 114 seconds
26 August 2012 - 13:30:23 Device: 0, finished
26 August 2012 - 13:30:23 Runtime: Device: 1, count: 0, 114 seconds
26 August 2012 - 13:30:23 Device: 1, finished
Ready ---------------------------------------------------------------------
Results:
Device: 0, device count: 1, average time / count: 114, average time on device: 114 Seconds (1 Minutes, 54 Seconds)
Device: 1, device count: 1, average time / count: 114, average time on device: 114 Seconds (1 Minutes, 54 Seconds)
Next ---------------------------------------------------------------------
26 August 2012 - 13:30:25 Start, devices: 2, device count: 2
26 August 2012 - 13:33:34 Runtime: Device: 1, count: 1, 184 seconds
26 August 2012 - 13:33:34 Device: 1, finished
26 August 2012 - 13:33:48 Runtime: Device: 1, count: 0, 198 seconds
26 August 2012 - 13:33:48 Device: 1, finished
26 August 2012 - 13:34:02 Runtime: Device: 0, count: 0, 212 seconds
26 August 2012 - 13:34:02 Device: 0, finished
26 August 2012 - 13:34:02 Runtime: Device: 0, count: 1, 212 seconds
26 August 2012 - 13:34:02 Device: 0, finished
Ready ---------------------------------------------------------------------
Results:
Device: 0, device count: 2, average time / count: 212, average time on device: 106 Seconds (1 Minutes, 46 Seconds)
Device: 1, device count: 2, average time / count: 191, average time on device: 95 Seconds (1 Minutes, 35 Seconds)
Next ---------------------------------------------------------------------
26 August 2012 - 13:34:03 Start, devices: 2, device count: 3
26 August 2012 - 13:34:49 Runtime: Device: 1, count: 2, 38 seconds
26 August 2012 - 13:34:49 ERROR: Device: 1, finished
26 August 2012 - 13:37:57 Runtime: Device: 1, count: 1, 226 seconds
26 August 2012 - 13:37:57 Device: 1, finished
26 August 2012 - 13:37:57 Runtime: Device: 1, count: 0, 226 seconds
26 August 2012 - 13:37:57 Device: 1, finished
26 August 2012 - 13:38:49 Runtime: Device: 0, count: 0, 278 seconds
26 August 2012 - 13:38:49 Device: 0, finished
26 August 2012 - 13:39:17 Runtime: Device: 0, count: 2, 306 seconds
26 August 2012 - 13:39:17 Device: 0, finished
26 August 2012 - 13:39:17 Runtime: Device: 0, count: 1, 306 seconds
26 August 2012 - 13:39:17 Device: 0, finished
Ready ---------------------------------------------------------------------
Results:
Device: 0, device count: 3, average time / count: 296, average time on device: 98 Seconds (1 Minutes, 38 Seconds)
Device: 1, device count: 3, average time / count: 163, average time on device: 54 Seconds (0 Minutes, 54 Seconds)
Next ---------------------------------------------------------------------
26 August 2012 - 13:39:19 Start, devices: 2, device count: 4
____________

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1275840 - Posted: 26 Aug 2012, 17:00:13 UTC - in response to Message 1275831.


Hi this is what you need?

Yep, just go the 41z running on a 590 as wel, I see a 4% increase.

Your last reading is highly suspicious, I expect the card to error out.

In 1.1 almost ready, I added an error check on the result.

You could check the count 3 and if you see the same difference investigate the stderr.txt to see if there is an overflow error.
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1275843 - Posted: 26 Aug 2012, 17:02:43 UTC - in response to Message 1275840.

Under some conditions, the program may generate a virus scanner warning.
I did sign the programs, if I didn't sign it, the program was shut down by my scanner almost at once.

____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

juan BFBProject donor
Volunteer tester
Avatar
Send message
Joined: 16 Mar 07
Posts: 5487
Credit: 316,277,715
RAC: 139,512
Brazil
Message 1275847 - Posted: 26 Aug 2012, 17:09:14 UTC - in response to Message 1275843.

The only thing is because the problem only apears on this host, in all the others i tests runs ok (all runs XP32 but with different GPUs combinations). Just report to you as a contribution in the development, even if is faster i can´t run more than 2WU at a time per GPU because the memory limitation on XP32.
____________

spitfire_mk_2
Avatar
Send message
Joined: 14 Apr 00
Posts: 463
Credit: 13,202,140
RAC: 5,881
United States
Message 1275852 - Posted: 26 Aug 2012, 17:25:39 UTC - in response to Message 1275613.

Well, I can't get it to work. I have one NVIDIA GT 520 installed. When I run the program with the default "Devices 1", "Count 0.50", I get this:

Starting test: (lunatics_x41g_win32_cuda32_perf.exe)
25 August 2012 - 20:56:50 Start, devices: 1, device count: 2
25 August 2012 - 20:57:07 Program failed to start on slot: 0
25 August 2012 - 20:57:07 Program failed to start on slot: 0
25 August 2012 - 20:57:08 Program failed to start on slot: 0
25 August 2012 - 20:57:08 Program failed to start on slot: 0

So ... what am I not doing correctly? (Yes, I stopped BOINC Manager before executing the performance utility).

My guess is that your card has enough video RAM to run only one task.

____________

Profile BMH
Avatar
Send message
Joined: 27 May 99
Posts: 333
Credit: 96,051,472
RAC: 25,118
United Kingdom
Message 1275873 - Posted: 26 Aug 2012, 18:00:44 UTC

Running the tool alongside GPU-Z gives a nice visual indication of which is best. Although running three tasks on my 460 test system showed slightly more solid GPU usage according to GPU-Z, 2 tasks (as expected) was the optimum.
____________
Brian.

Horacio
Send message
Joined: 14 Jan 00
Posts: 536
Credit: 75,956,386
RAC: 4,833
Argentina
Message 1275901 - Posted: 26 Aug 2012, 19:05:59 UTC - in response to Message 1275433.


Any chance of adding 5 & 6 ?

You can do that manually. (manual button)
On auto mode the program tens to hang after 4 for sure. Probably out of memory, but doesn't exit.

I could add a max in the config.xml.


First, Thanks Fred, another really good tool!

Now, If I set the count to 0.20 for a manual test it doesnt start... It shows a message saying that only valus of 1, 0.5, 0.33 and 0.25 are valid...
May be instead of this message you could give a warning saying that it could fail with more than 4 instances per card giving the option to do it anyway or not on user choice...

Another suggestion is to anchor the controls so they get resized/relocated when the users resizes the main window... Its just cosmetics anyway... (And I know that while this it's easy on .NET, it may be a lot of work if you are using the native SDK and C++)
____________

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1276220 - Posted: 27 Aug 2012, 14:15:00 UTC - in response to Message 1275901.
Last modified: 27 Aug 2012, 14:15:29 UTC

The list so far, I hope there is no typo.

As you can see the 590 still leads.
But with Jason working on the 6xx, this will change soon.

x41g

460 158 Sec / Wu Oc
560 150 Sec / Wu
295 130 Sec / Wu
670 129 Sec / Wu
680 107 Sec / Wu
480 100 Sec / Wu
690 58 Sec / Wu
590 53 Sec / Wu

x41z

295 Not running stable
480 89 Sec / Wu
680 75 Sec / Wu cuda 5
590 53 Sec / Wu
690 44 Sec / Wu
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

juan BFBProject donor
Volunteer tester
Avatar
Send message
Joined: 16 Mar 07
Posts: 5487
Credit: 316,277,715
RAC: 139,512
Brazil
Message 1276221 - Posted: 27 Aug 2012, 14:26:45 UTC - in response to Message 1276220.
Last modified: 27 Aug 2012, 14:28:22 UTC

I´m not shure but i think i see in some forum some special build for the 295 with x41z, maybe you could ask Jason about that.

About the table, ok i could see lower is better, but what those number realy means? Just curiosity.
____________

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1276222 - Posted: 27 Aug 2012, 14:31:36 UTC - in response to Message 1276221.

I´m not shure but i think i see in some forum some special build for the 295 with x41z, maybe you could ask Jason about that.

About the table, ok i could see lower is better, but what those number realy means? Just curiosity.


There is, a cuda 3.2 version, but it errors out -9 overflow on one of them.
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1276223 - Posted: 27 Aug 2012, 14:35:11 UTC - in response to Message 1275901.


Another suggestion is to anchor the controls so they get resized/relocated when the users resizes the main window... Its just cosmetics anyway... (And I know that while this it's easy on .NET, it may be a lot of work if you are using the native SDK and C++)

I see, the resize is still on.
Will move the buttons to the top, that makes the text resizable.

You could move the buttons around as well, but very easy for something to mess things up.

And on 125% and 150% this cause other problems.

____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

Russ Sheldon
Avatar
Send message
Joined: 18 Jan 01
Posts: 381
Credit: 476,504,169
RAC: 207,505
United States
Message 1276275 - Posted: 27 Aug 2012, 16:27:01 UTC - in response to Message 1276223.
Last modified: 27 Aug 2012, 17:24:45 UTC

Thanks Fred! Runs great with _g. Won't run with _z. Program failed to start on slot: 0
____________
“Grandpa, I heard a story about a guy named Russ who crunched over 700,000 RAC for an unappreciative SETI in Oct. 2012. Is that legend true?”
“Yes Billy, and some people say he still crunches. All we can do is watch the skies... and hope.”

Horacio
Send message
Joined: 14 Jan 00
Posts: 536
Credit: 75,956,386
RAC: 4,833
Argentina
Message 1276279 - Posted: 27 Aug 2012, 16:39:27 UTC
Last modified: 27 Aug 2012, 16:39:46 UTC

This are my results:

Both tests on Win 7 Pro 32b.
I7-3770 (3.5Ghz constant with turbo boosted) on z68 chipset. BOINC full stopped.
Driver 301.42


GTX680 Default Clock (core at 1111Mhz -full boost-, Mem at 3004Mhz) Pcie3.016x
Device: 0, device count: 1, average time / count: 170, average time on device: 170 Seconds (2 Minutes, 50 Seconds)
Device: 0, device count: 2, average time / count: 244, average time on device: 122 Seconds (2 Minutes, 2 Seconds)
Device: 0, device count: 3, average time / count: 355, average time on device: 118 Seconds (1 Minutes, 58 Seconds)
Device: 0, device count: 4, average time / count: 477, average time on device: 119 Seconds (1 Minutes, 59 Seconds)
The best average time found: 118 Seconds (1 Minutes, 58 Seconds), with count: 0.33 (3)


GTX680 OC'ed (core at 1215Mhz -full boost-, Mem at 3105Mhz) Pcie3.016x
Device: 0, device count: 1, average time / count: 163, average time on device: 163 Seconds (2 Minutes, 43 Seconds)
Device: 0, device count: 2, average time / count: 235, average time on device: 117 Seconds (1 Minutes, 57 Seconds)
Device: 0, device count: 3, average time / count: 340, average time on device: 113 Seconds (1 Minutes, 53 Seconds)
Device: 0, device count: 4, average time / count: 456, average time on device: 114 Seconds (1 Minutes, 54 Seconds)
The best average time found: 113 Seconds (1 Minutes, 53 Seconds), with count: 0.33 (3)

Weird thing... my gtx680 seems to be slower than others, even OC'ed... And nothing else running on the host other than MSI afterburner and Process Laso to set the apps at normal priority...
____________

Profile S@NL - eFMer - efmer.com/boincProject donor
Volunteer tester
Avatar
Send message
Joined: 7 Jun 99
Posts: 512
Credit: 131,350,280
RAC: 26,458
United States
Message 1276313 - Posted: 27 Aug 2012, 17:32:50 UTC - in response to Message 1276279.


Weird thing... my gtx680 seems to be slower than others, even OC'ed... And nothing else running on the host other than MSI afterburner and Process Laso to set the apps at normal priority...

Yep seems to be on the low side.
NVIDIA 3d setting->power management->max performance.

What I've seen so far is that the 6xx series is slower than the 4xx and 5xx series.
Probably something to do with optimizing.
____________
TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking.

Previous · 1 · 2 · 3 · 4 · 5 . . . 13 · Next

Message boards : Number crunching : Optimize your GPU. Find the value the easy way.

Copyright © 2014 University of California