View Full Version : Einstein@Home - Binary Radio Pulsar Search (Arecibo) ???
Does anybody know if this is a CPU Wu or a GPU Wu, I'm at 9800 Hr's on it & haven't seen any work for it in a long time even though I have my Einstein Preferences set to a High Resource Value ... Thanks
On Windows machines this runs on Intel GPU. Good luck!
zombie67
10-20-14, 08:58 AM
Does anybody know if this is a CPU Wu or a GPU Wu, I'm at 9800 Hr's on it & haven't seen any work for it in a long time even though I have my Einstein Preferences set to a High Resource Value ... Thanks
There are two Arecibo apps, one for GPU and one for CPU. http://einstein.phys.uwm.edu/apps.php
Binary Radio Pulsar Search (Arecibo, GPU)
Binary Radio Pulsar Search (Arecibo)
I tried to check the server status to see if there was any work for the CPU version of that app. Of course they name the apps something completely different on the server status page (why do projects do that?), so it's impossible to decode.
So then I tried getting some work on a windows machine. The neat thing about Einstein, it will show you the server log URL in your even log. This is what mine shows:
http://einstein5.aei.uni-hannover.de/EinsteinAtHome/host_sched_logs/11662/11662309
2014-10-20 12:46:18.6062 [PID=21280] Request: [USER#xxxxx] [HOST#11662309] [IP xxx.xxx.xxx.152] client 7.2.42
2014-10-20 12:46:18.6066 [PID=21280] [debug] [HOST#11662309] Resetting nresults_today
2014-10-20 12:46:18.6067 [PID=21280] [send] effective_ncpus 5 max_jobs_on_host_cpu 999999 max_jobs_on_host 999999
2014-10-20 12:46:18.6067 [PID=21280] [send] effective_ngpus 0 max_jobs_on_host_gpu 999999
2014-10-20 12:46:18.6068 [PID=21280] [send] Not using matchmaker scheduling; Not using EDF sim
2014-10-20 12:46:18.6068 [PID=21280] [send] CPU: req 561600.00 sec, 5.00 instances; est delay 0.00
2014-10-20 12:46:18.6068 [PID=21280] [send] work_req_seconds: 561600.00 secs
2014-10-20 12:46:18.6068 [PID=21280] [send] available disk 73.12 GB, work_buf_min 86400
2014-10-20 12:46:18.6068 [PID=21280] [send] active_frac 0.990117 on_frac 0.914468 DCF 1.606610
2014-10-20 12:46:18.6074 [PID=21280] [send] [HOST#11662309] is reliable
2014-10-20 12:46:18.6076 [PID=21280] [send] set_trust: random choice for error rate 0.000727: yes
2014-10-20 12:46:18.6076 [PID=21280] [mixed] sending non-locality work first (0.4210)
2014-10-20 12:46:18.6265 [PID=21280] [send] [HOST#11662309] will accept beta work. Scanning for beta work.
2014-10-20 12:46:18.6469 [PID=21280] [version] Checking plan class 'opencl-intel_gpu'
2014-10-20 12:46:18.6487 [PID=21280] [version] reading plan classes from file '/BOINC/projects/EinsteinAtHome/plan_class_spec.xml'
2014-10-20 12:46:18.6487 [PID=21280] [version] numerical Windows version: 601760100 (Microsoft Windows 7 Professional x64 Edition, Service Pack 1, (06.01.7601.00))
2014-10-20 12:46:18.6488 [PID=21280] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 12:46:18.6488 [PID=21280] [version] No Intel GPU devices found
2014-10-20 12:46:18.6488 [PID=21280] [version] Checking plan class 'opencl-intel_gpu'
2014-10-20 12:46:18.6489 [PID=21280] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 12:46:18.6489 [PID=21280] [version] No Intel GPU devices found
2014-10-20 12:46:18.6489 [PID=21280] [version] no app version available: APP#19 (einsteinbinary_BRP4) PLATFORM#9 (windows_x86_64) min_version 0
2014-10-20 12:46:18.6489 [PID=21280] [version] no app version available: APP#19 (einsteinbinary_BRP4) PLATFORM#2 (windows_intelx86) min_version 0
2014-10-20 12:46:18.6583 [PID=21280] [mixed] sending locality work second
2014-10-20 12:46:18.6604 [PID=21280] [debug] [HOST#11662309] MSG(high) No work sent
2014-10-20 12:46:18.6604 [PID=21280] [debug] [HOST#11662309] MSG(high) see scheduler log messages on http://einstein5.aei.uni-hannover.de/EinsteinAtHome/host_sched_logs/11662/11662309
2014-10-20 12:46:18.6604 [PID=21280] Sending reply to [HOST#11662309]: 0 results, delay req 60.00
2014-10-20 12:46:18.6614 [PID=21280] Scheduler ran 0.058 seconds
it looks like something is messed up. My machine is asking for CPU work, but then the server returns with "No Intel GPU devices found".
And that leads me back to the apps page. The two apps for windows are:
Windows/x86 1.34 (opencl-intel_gpu) 2 Sep 2013 13:37:10 UTC
Microsoft Windows running on an AMD x86_64 or Intel EM64T CPU 1.34 (opencl-intel_gpu) 9 Jul 2013 8:28:34 UTC
They made the apps OpenCL. That SHOULD be okay, as OpenCL is supposed to be able to run on both CPUs and GPUs. But the server thinks it is only for GPUs for some reason.
Looks like you will need to post something in their forums to see what the problem is. Maybe they aren't even aware.
Here are the BRPS apps with corresponding server names:
Binary Radio Pulsar Search (Arecibo)
BRP4
Binary Radio Pulsar Search (Arecibo, GPU)
BRP4G
Binary Radio Pulsar Search (Perseus Arm Survey)
BRP5
I'm currently crunching Binary Radio Pulsar Search (Arecibo) on an Intel GPU. As far as I know, there are no BRPS CPU apps.
zombie67
10-20-14, 11:45 AM
I'm currently crunching Binary Radio Pulsar Search (Arecibo) on an Intel GPU. As far as I know, there are no BRPS CPU apps.
The app at the bottom of the apps page (http://einstein.phys.uwm.edu/apps.php) is for CPUs only. The two apps for windows are OpenCL, which should be fine as OpenCL runs on CPUs as well as GPUs. But their server is messed up, I believe. It thinks the client must have an OpenCL GPU, when it really does not.
I'm calling for Intel GPU Work now but haven't received any yet, all I need to do is run a few day's to get to 10,000 Hr's on that App ...
zombie67
10-20-14, 12:20 PM
I'm calling for Intel GPU Work now but haven't received any yet, all I need to do is run a few day's to get to 10,000 Hr's on that App ...
Just checking, do you have a CPU with ivy bridge (HD 4000) or later? Also, do you have the intel OpenCL GPU driver installed? Finally, does BOINC see your Intel GPU (HD 4000 or later) in the event log when it starts?
This is what mine looks like, I have Intel GPU selected in project preferences and I'm getting work.
10/20/2014 11:41:16 AM | | Starting BOINC client version 7.2.42 for windows_x86_64
10/20/2014 11:41:16 AM | | log flags: file_xfer, sched_ops, task
10/20/2014 11:41:16 AM | | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
10/20/2014 11:41:16 AM | | Data directory: C:\ProgramData\BOINC
10/20/2014 11:41:16 AM | | Running under account DLewis
10/20/2014 11:41:16 AM | | CUDA: NVIDIA GPU 0: Quadro K2000M (driver version 333.11, CUDA version 6.0, compute capability 3.0, 2048MB, 1978MB available, 572 GFLOPS peak)
10/20/2014 11:41:16 AM | | OpenCL: NVIDIA GPU 0: Quadro K2000M (driver version 333.11, device version OpenCL 1.1 CUDA, 2048MB, 1978MB available, 572 GFLOPS peak)
10/20/2014 11:41:16 AM | | OpenCL: Intel GPU 0: Intel(R) HD Graphics 4000 (driver version 9.18.10.3165, device version OpenCL 1.2, 1624MB, 1624MB available, 45 GFLOPS peak)
10/20/2014 11:41:16 AM | | OpenCL CPU: Intel(R) Core(TM) i7-3840QM CPU @ 2.80GHz (OpenCL driver vendor: Intel(R) Corporation, driver version 1.2, device version OpenCL 1.2 (Build 63463))
Mine says:
PBOYZTOY01
6 10/20/2014 1:25:54 PM OpenCL: Intel GPU 0: Intel(R) HD Graphics 4000 (driver version 8.15.10.2761, device version OpenCL 1.1, 1496MB, 1496MB available, 45 GFLOPS peak)
7 10/20/2014 1:25:54 PM OpenCL CPU: Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz (OpenCL driver vendor: Intel(R) Corporation, driver version 1.1, device version OpenCL 1.1 (Build 31360.31441))
Maybe I don't have a late enough version of the Intel Open CL ?
Here's the Scheduler Log:
2014-10-20 17:37:13.4959 [PID=25946] Request: [USER#xxxxx] [HOST#11634490] [IP xxx.xxx.xxx.206] client 7.4.12
2014-10-20 17:37:13.4964 [PID=25946] [send] effective_ncpus 8 max_jobs_on_host_cpu 999999 max_jobs_on_host 999999
2014-10-20 17:37:13.4964 [PID=25946] [send] effective_ngpus 1 max_jobs_on_host_gpu 999999
2014-10-20 17:37:13.4964 [PID=25946] [send] Not using matchmaker scheduling; Not using EDF sim
2014-10-20 17:37:13.4964 [PID=25946] [send] CPU: req 2841.71 sec, 8.00 instances; est delay 0.00
2014-10-20 17:37:13.4964 [PID=25946] [send] Intel GPU: req 194400.00 sec, 1.00 instances; est delay 0.00
2014-10-20 17:37:13.4964 [PID=25946] [send] work_req_seconds: 2841.71 secs
2014-10-20 17:37:13.4964 [PID=25946] [send] available disk 10.48 GB, work_buf_min 172800
2014-10-20 17:37:13.4964 [PID=25946] [send] active_frac 1.000000 on_frac 0.998982 DCF 1.000000
2014-10-20 17:37:13.4972 [PID=25946] [send] [HOST#11634490] not reliable; max_result_day 31
2014-10-20 17:37:13.4973 [PID=25946] [send] set_trust: error rate 0.100000 > 0.050000, don't trust
2014-10-20 17:37:13.4973 [PID=25946] [mixed] sending non-locality work first (0.0231)
2014-10-20 17:37:13.4973 [PID=25946] [send] [HOST#11634490] will accept beta work. Scanning for beta work.
2014-10-20 17:37:13.5268 [PID=25946] [version] Checking plan class 'BRP5-opencl-ati'
2014-10-20 17:37:13.5287 [PID=25946] [version] reading plan classes from file '/BOINC/projects/EinsteinAtHome/plan_class_spec.xml'
2014-10-20 17:37:13.5287 [PID=25946] [version] numerical Windows version: 601760100 (Microsoft Windows 7 Professional x64 Edition, Service Pack 1, (06.01.7601.00))
2014-10-20 17:37:13.5287 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5287 [PID=25946] [version] No ATI devices found
2014-10-20 17:37:13.5288 [PID=25946] [version] Checking plan class 'BRP5-cuda32'
2014-10-20 17:37:13.5288 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5288 [PID=25946] [version] No CUDA devices found
2014-10-20 17:37:13.5288 [PID=25946] [version] Checking plan class 'BRP5-cuda32-nv301'
2014-10-20 17:37:13.5288 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5288 [PID=25946] [version] No CUDA devices found
2014-10-20 17:37:13.5288 [PID=25946] [version] Checking plan class 'BRP5-opencl-ati'
2014-10-20 17:37:13.5288 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5288 [PID=25946] [version] No ATI devices found
2014-10-20 17:37:13.5288 [PID=25946] [version] no app version available: APP#23 (einsteinbinary_BRP5) PLATFORM#9 (windows_x86_64) min_version 0
2014-10-20 17:37:13.5288 [PID=25946] [version] no app version available: APP#23 (einsteinbinary_BRP5) PLATFORM#2 (windows_intelx86) min_version 0
2014-10-20 17:37:13.5289 [PID=25946] [version] Checking plan class 'BRP4G-opencl-ati'
2014-10-20 17:37:13.5289 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5289 [PID=25946] [version] No ATI devices found
2014-10-20 17:37:13.5289 [PID=25946] [version] Checking plan class 'BRP4G-cuda32'
2014-10-20 17:37:13.5289 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5289 [PID=25946] [version] No CUDA devices found
2014-10-20 17:37:13.5289 [PID=25946] [version] Checking plan class 'BRP4G-cuda32-nv301'
2014-10-20 17:37:13.5289 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5290 [PID=25946] [version] No CUDA devices found
2014-10-20 17:37:13.5290 [PID=25946] [version] Checking plan class 'BRP4G-opencl-ati'
2014-10-20 17:37:13.5290 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5290 [PID=25946] [version] No ATI devices found
2014-10-20 17:37:13.5290 [PID=25946] [version] no app version available: APP#25 (einsteinbinary_BRP4G) PLATFORM#9 (windows_x86_64) min_version 0
2014-10-20 17:37:13.5290 [PID=25946] [version] no app version available: APP#25 (einsteinbinary_BRP4G) PLATFORM#2 (windows_intelx86) min_version 0
2014-10-20 17:37:13.5293 [PID=25946] [version] Checking plan class 'opencl-intel_gpu'
2014-10-20 17:37:13.5294 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5294 [PID=25946] [version] OpenCL device version required min: 102, supplied: 101
2014-10-20 17:37:13.5294 [PID=25946] [version] Checking plan class 'opencl-intel_gpu'
2014-10-20 17:37:13.5294 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5294 [PID=25946] [version] OpenCL device version required min: 102, supplied: 101
2014-10-20 17:37:13.5295 [PID=25946] [version] no app version available: APP#19 (einsteinbinary_BRP4) PLATFORM#9 (windows_x86_64) min_version 0
2014-10-20 17:37:13.5295 [PID=25946] [version] no app version available: APP#19 (einsteinbinary_BRP4) PLATFORM#2 (windows_intelx86) min_version 0
2014-10-20 17:37:13.5392 [PID=25946] [mixed] sending locality work second
2014-10-20 17:37:13.5412 [PID=25946] [debug] [HOST#11634490] MSG(high) No work sent
2014-10-20 17:37:13.5413 [PID=25946] [debug] [HOST#11634490] MSG(high) No work is available for Gamma-ray pulsar search #3
2014-10-20 17:37:13.5413 [PID=25946] [debug] [HOST#11634490] MSG(high) see scheduler log messages on http://einstein5.aei.uni-hannover.de/EinsteinAtHome/host_sched_logs/11634/11634490
2014-10-20 17:37:13.5413 [PID=25946] Sending reply to [HOST#11634490]: 0 results, delay req 60.00
2014-10-20 17:37:13.5423 [PID=25946] Scheduler ran 0.050 seconds
I'd start with a driver update. Turns out mine are outdated as well... just not quite as old as yours!
Update: An OpenCL 1.2 driver is required: http://einstein.phys.uwm.edu/forum_thread.php?id=10215
2014-10-20 17:37:13.5294 [PID=25946] [version] OpenCL device version required min: 102, supplied: 101
Mine says:
2014-10-20 17:37:13.5294 [PID=25946] [version] OpenCL device version required min: 102, supplied: 101
2014-10-20 17:37:13.5294 [PID=25946] [version] Checking plan class 'opencl-intel_gpu'
2014-10-20 17:37:13.5294 [PID=25946] [version] parsed project prefs setting 'gpu_util_brp': 1.000000
2014-10-20 17:37:13.5294 [PID=25946] [version] OpenCL device version required min: 102, supplied: 101
Yup, I think you are correct that you need a driver update.
Well it was like Pulling Pickers from my Pecker but I managed to get the Drivers Updated on 3 Box's & am running the Einstein Intel Wu's on them, all I need is about 4 Day's running & I'l be over 10k Hr's for the (Arecibo) Wu's ...
I have one android running around the clock on Binary Radio Pulsar Search (Arecibo). It's not much credit, but it's something and helps with WUProp.
Its the Gravitational Wave S6 Directed Search (CasA) a that is rare getting work for on other machines and the Gamma-ray pulsar search #2 and Gravitational Wave S6 LineVeto search (extended) that I think are done.
Powered by vBulletin® Version 4.2.4 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.