How can I use 2 gpu's at the same time? ATI + Nvidia?

Totally
Totally
Joined: 3 May 13
Posts: 6
Credit: 4363434
RAC: 0
Topic 196938

How can I use 2 gpu's at the same time? ATI + Nvidia? use_all_gpus_1 is in the config.xml. Both applications have received wu's but it is computed only on the Nvidia.

Alex
Alex
Joined: 1 Mar 05
Posts: 451
Credit: 507424926
RAC: 77287

How can I use 2 gpu's at the same time? ATI + Nvidia?

Quote:
How can I use 2 gpu's at the same time? ATI + Nvidia? use_all_gpus_1 is in the config.xml. Both applications have received wu's but it is computed only on the Nvidia.


Should not be a problem, I have two pc's doing that.
Can you please post the first lines of the messages; there should be messages about your gpu's abilities and detection.
Did you enable both gpu's in your personal settings?

Totally
Totally
Joined: 3 May 13
Posts: 6
Credit: 4363434
RAC: 0

Use ATI and Nvidia are

Use ATI and Nvidia are enable,

03.05.2013 14:55:43 | | Starting BOINC client version 7.0.59 for windows_x86_64
03.05.2013 14:55:43 | | log flags: file_xfer, sched_ops, task
03.05.2013 14:55:43 | | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
03.05.2013 14:55:43 | | Data directory: C:\ProgramData\BOINC
03.05.2013 14:55:43 | | Running under account Totally
03.05.2013 14:55:43 | | Processor: 8 GenuineIntel Intel(R) Xeon(R) CPU E3-1230 V2 @ 3.30GHz [Family 6 Model 58 Stepping 9]
03.05.2013 14:55:43 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 popcnt aes syscall nx lm vmx smx tm2 pbe
03.05.2013 14:55:43 | | OS: Microsoft Windows 7: Enterprise x64 Edition, Service Pack 1, (06.01.7601.00)
03.05.2013 14:55:43 | | Memory: 15.94 GB physical, 19.94 GB virtual
03.05.2013 14:55:43 | | Disk: 90.00 GB total, 23.88 GB free
03.05.2013 14:55:43 | | Local time is UTC +2 hours
03.05.2013 14:55:43 | | VirtualBox version: 4.2.12
03.05.2013 14:55:43 | | CUDA: NVIDIA GPU 0: GeForce GTS 450 (driver version 314.22, CUDA version 5.0, compute capability 2.1, 1024MB, 962MB available, 601 GFLOPS peak)
03.05.2013 14:55:43 | | CAL: ATI GPU 0: AMD Radeon HD 7700 series (Cape Verde) (CAL version 1.4.1741, 1024MB, 984MB available, 2048 GFLOPS peak)
03.05.2013 14:55:43 | | OpenCL: NVIDIA GPU 0: GeForce GTS 450 (driver version 314.22, device version OpenCL 1.1 CUDA, 1024MB, 962MB available, 601 GFLOPS peak)
03.05.2013 14:55:43 | | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 7700 series (Cape Verde) (driver version 1084.4 (VM), device version OpenCL 1.2 AMD-APP (1084.4), 1024MB, 984MB available, 2048 GFLOPS peak)

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5893653
RAC: 51

In your account, project

In your account, project preferences, what are the entries for:

Use ATI GPU
Enforced by version 6.10+
Use NVIDIA GPU
Enforced by version 6.10+

Run only the selected applications
- Binary Radio Pulsar Search (Arecibo)
- Binary Radio Pulsar Search (Perseus Arm Survey)
- Gravitational Wave S6 LineVeto search (extended)
- Gamma-ray pulsar search #2

??

Totally
Totally
Joined: 3 May 13
Posts: 6
Credit: 4363434
RAC: 0

Use ATI GPU Enforced by

Use ATI GPU
Enforced by version 6.10+ ja
Use NVIDIA GPU
Enforced by version 6.10+ ja

Run only the selected applications
Binary Radio Pulsar Search (Arecibo): ja
Binary Radio Pulsar Search (Perseus Arm Survey): ja
Gravitational Wave S6 LineVeto search (extended): nein
Gamma-ray pulsar search #2: nein

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5893653
RAC: 51

Any specific reason why

Any specific reason why you're using beta-7.0.59 and not the recommended 7.0.64?
In BOINC Manager, what's under Activity?
Do you have any work request that asks for work for CPU, ATI GPU and Nvidia GPU? (check through stdoutdae.txt in your BOINC data directory (Data directory: C:\ProgramData\BOINC))

In BOINC Manager, Projects tab, select Einstein@Home, click Properties. Does it say AMD/ATI GPU work fetch deferred for xxx? If so, by how much?
What then happens when you do BOINC Manager->Advanced->Do network communication?

Totally
Totally
Joined: 3 May 13
Posts: 6
Credit: 4363434
RAC: 0

Update to 7.0.64 is done,

Update to 7.0.64 is done, there was no particular reason for the beta.
Under active 6 x WGC, 2 x Einstein (now Nvidia, ATI previously) and 1 x WuProb. No other projects are activ and Einstein the only on gpu.

03.05.2013 19:40:31 | Einstein@Home | Requesting new tasks for NVIDIA and ATI
03.05.2013 19:40:33 | Einstein@Home | Scheduler request completed: got 0 new tasks
03.05.2013 19:40:33 | Einstein@Home | No work sent
03.05.2013 19:40:33 | Einstein@Home | No work is available for Binary Radio Pulsar Search (Arecibo)
03.05.2013 19:40:33 | Einstein@Home | No work is available for Binary Radio Pulsar Search (Perseus Arm Survey)
03.05.2013 19:40:33 | Einstein@Home | (reached daily quota of 512 tasks)
03.05.2013 19:40:33 | Einstein@Home | Project has no jobs available

The reason for no new jobs, I have reset the project

AMD/ATI GPU work fetch deferred 10:00
NVIDIA GPU work fetch deferred 10:00

Do network communication ->
03.05.2013 19:48:57 | boincsimap | Fetching scheduler list
03.05.2013 19:48:58 | World Community Grid | Temporarily failed upload of E213094_163_C.33.C29H19N3Si.00393182.3.set1d06_0_1: connect() failed
03.05.2013 19:48:58 | World Community Grid | Backing off 4 hr 27 min 13 sec on upload of E213094_163_C.33.C29H19N3Si.00393182.3.set1d06_0_1
03.05.2013 19:49:11 | | Project communication failed: attempting access to reference site
03.05.2013 19:49:12 | | Internet access OK - project servers may be temporarily down.
03.05.2013 19:50:05 | Einstein@Home | Sending scheduler request: To fetch work.
03.05.2013 19:50:05 | Einstein@Home | Requesting new tasks for ATI
03.05.2013 19:50:07 | Einstein@Home | Scheduler request completed: got 0 new tasks
03.05.2013 19:50:07 | Einstein@Home | No work sent
03.05.2013 19:50:07 | Einstein@Home | No work is available for Binary Radio Pulsar Search (Arecibo)
03.05.2013 19:50:07 | Einstein@Home | No work is available for Binary Radio Pulsar Search (Perseus Arm Survey)
03.05.2013 19:50:07 | Einstein@Home | (reached daily quota of 512 tasks)
03.05.2013 19:50:07 | Einstein@Home | Project has no jobs available

SIMAP and WGC blocked for the Pentathlon, this is okay.

My app_config


einsteinbinary_BRP4
4

0.5
0.5


MAGIC Quantum Mechanic
MAGIC Quantum M...
Joined: 18 Jan 05
Posts: 1892
Credit: 1415247770
RAC: 1169092

It appears that yours is

It appears that yours is working now with both cards.

http://einsteinathome.org/host/7174012/tasks

http://einsteinathome.org/host/7174012/tasks&offset=0&show_names=1&state=2&appid=0

9 finished so far and they are all still pending which is why you don't see credit yet.

And 117 in progress.

Horacio
Horacio
Joined: 3 Oct 11
Posts: 205
Credit: 80557243
RAC: 0

You are not getting new tasks

You are not getting new tasks because you have already reached the daily quota...
This quota is low right now probably because you aborted several tasks recently, it's supossed that it will rise by itself when your host returns the work and it is validated...
Everything else seems to be Ok, indeed, in your tasks page there are BRP WUs allocated (in progress) for both Nvidia a ATI GPUs, and there are also WUs returned from both GPUs (pending), so, you should see both GPUs beeing used... and with your settings you should see 4 BRP tasks beeing crunched (2 in each GPU)...

Is the BOINC manager showing something different? What shows?

Totally
Totally
Joined: 3 May 13
Posts: 6
Credit: 4363434
RAC: 0

RE: It appears that yours

Quote:

It appears that yours is working now with both cards.

http://einsteinathome.org/host/7174012/tasks

http://einsteinathome.org/host/7174012/tasks&offset=0&show_names=1&state=2&appid=0

9 finished so far and they are all still pending which is why you don't see credit yet.

And 117 in progress.

Quote:

You are not getting new tasks because you have already reached the daily quota...
This quota is low right now probably because you aborted several tasks recently, it's supossed that it will rise by itself when your host returns the work and it is validated...
Everything else seems to be Ok, indeed, in your tasks page there are BRP WUs allocated (in progress) for both Nvidia a ATI GPUs, and there are also WUs returned from both GPUs (pending), so, you should see both GPUs beeing used... and with your settings you should see 4 BRP tasks beeing crunched (2 in each GPU)...

Is the BOINC manager showing something different? What shows?

The reason for results from both cards are that, when the manager start they use the ati and when finished the wu's the nvidia is used and no ati anymore.
This means, there are only 2 tasks active, nvidia or ati but not all 4 together and this is my problem.

Horacio
Horacio
Joined: 3 Oct 11
Posts: 205
Credit: 80557243
RAC: 0

RE: The reason for results

Quote:
The reason for results from both cards are that, when the manager start they use the ati and when finished the wu's the nvidia is used and no ati anymore.
This means, there are only 2 tasks active, nvidia or ati but not all 4 together and this is my problem.


Ok.
I guess that this would need to set some debug options to see why the BOINC client is doing that... But that's something out of my expertise...

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.