1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Guide to running multiple GPU work units

Discussion in 'World Community Grid (WCG)' started by KieX, Oct 21, 2012.

  1. KieX

    Joined:
    Dec 18, 2009
    Messages:
    2,408 (1.36/day)
    Thanks Received:
    2,461
    Location:
    London, UK
    This level of variability is a very recent phenomena with the GPU WU though. WCG used to be incredibly simple and straight forward. :ohwell: But they're a very collaborative bunch, so hopefully we'll see even these things simplified to that standard.
     
  2. BUCK NASTY

    BUCK NASTY F@H Mod & 4P Enthusiust Staff Member

    Joined:
    Aug 8, 2007
    Messages:
    4,447 (1.69/day)
    Thanks Received:
    3,533
    Location:
    Tallahassee, FL. U.S.A.
    Yeah, I thought F@H was the "difficult to set-up" DC project.:laugh: I gave up on multiple work units in favor of consistency.
     
    KieX says thanks.
  3. Jstn7477

    Jstn7477

    Joined:
    Aug 30, 2009
    Messages:
    3,881 (2.06/day)
    Thanks Received:
    1,560
    Location:
    Sarasota, Florida, USA
    Got it working with my GTX 470 and GTX 465 in my FX-8150 system, thanks!
     
    KieX says thanks.
    Crunching for Team TPU More than 25k PPD
  4. KieX

    Joined:
    Dec 18, 2009
    Messages:
    2,408 (1.36/day)
    Thanks Received:
    2,461
    Location:
    London, UK
    Awesome :rockout:
     
  5. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    So can someone give me the app_info file that works for HCC CPU and HCC GPU on AMD cards w/ the newest 7.05 WUs?
     
    Crunching for Team TPU
  6. KieX

    Joined:
    Dec 18, 2009
    Messages:
    2,408 (1.36/day)
    Thanks Received:
    2,461
    Location:
    London, UK
    Here you go. Make sure to change the coprocessor count, CPU is set to use full cores. Will add to the OP too.
     

    Attached Files:

    stinger608, Daimus and [Ion] say thanks.
  7. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    Thanks a ton! :)

    My cache will be empty in a few hours, so I'll do this after class :)
     
    KieX says thanks.
    Crunching for Team TPU
  8. mstenholm

    Joined:
    Nov 7, 2009
    Messages:
    2,069 (1.14/day)
    Thanks Received:
    1,172
    On one of mine PCs (i7-920/7770/GPU only, 3 GPU at the time, Ave Turnaround (tbf) 1,608.58) with version 7.0.28 and 3 days buffer I have anywhere from 1 to 12 GPU WU ready to start at any given time). A few hours is pure luxury :) On top of that my internet is on and off these days. I can't wait until I install my new 7950 tomorrow....:twitch:
     
    KieX says thanks.
    Crunching for Team TPU More than 25k PPD
  9. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    This was just running with no app_info file (and it has a few hours of CPU WUs left--the GPU is already idle).

    We'll see how things go w/ the app_info and whether it will keep pulling down new WUs.
     
    Crunching for Team TPU
  10. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    Well, I think I'd call this a success. I was running 1 WU at a time, and it was completing in about 5.5 or 6 minutes (HD7770). Now I'm running 3 at once, and they're taking about 9:30 each :toast:

    Now to make sure that it's going to keep getting new work as it submits the finished stuff--I'll keep a close eye on it today and see :toast:
     
    KieX says thanks.
    Crunching for Team TPU
  11. brandonwh64

    brandonwh64 Addicted to Bacon and StarCrunches!!!

    Joined:
    Sep 6, 2009
    Messages:
    18,680 (9.97/day)
    Thanks Received:
    6,154
    Location:
    Chatsworth, GA
    All of the .XML files I have tried throw errors now on both my cruncher at work and at home :/ I guess its back to using the 2600K only and one GPU
     
    Crunching for Team TPU
  12. mstenholm

    Joined:
    Nov 7, 2009
    Messages:
    2,069 (1.14/day)
    Thanks Received:
    1,172
    With the new HCC ones my average is 6:55 (timed to be around 9 sec longer then recorded) with three. 3.5 GHz, slow HD, no CPU units, 3 at the time. I didn't time the rig that runs 4 at the time but my estimate is just less then 9 min so no big advantage in going up in numbers on a 7770. Both figures is without running GPU tasks in the background (games/video/general surfing). Is your 9:30 achieved during normal PC use?

    I will do the same exercise (3, 4, 5 and 6 GPU units) with my new 7950 tomorrow on the same 3.5 Ghz CPU. I know that the card can run 10 and even 12 at the time (Source: KieX) but I like to keep my temperatures below 55 C and my office is just next to the room where my wood stove is burning.
     
    KieX, Norton and [Ion] say thanks.
    Crunching for Team TPU More than 25k PPD
  13. Jstn7477

    Jstn7477

    Joined:
    Aug 30, 2009
    Messages:
    3,881 (2.06/day)
    Thanks Received:
    1,560
    Location:
    Sarasota, Florida, USA
    EDIT: Going to try 7.0.38 and see if it helps.

    EDIT2: Still doesn't work. I can't believe BOINC will only use one GPU if it is "superior than the other." You would think they would want the most output possible???

    So I put my 7770 in the same machine as my 7950 and the 7770 is reporting as "not used" when I specified <use_all_gpus>1</use_all_gpus> in cc_config.xml. Any ideas?

    Code:
    11/14/2012 9:55:32 PM |  | Starting BOINC client version 7.0.28 for windows_x86_64
    11/14/2012 9:55:32 PM |  | log flags: file_xfer, sched_ops, task
    11/14/2012 9:55:32 PM |  | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
    11/14/2012 9:55:32 PM |  | Data directory: C:\ProgramData\BOINC
    11/14/2012 9:55:32 PM |  | Running under account Justin
    11/14/2012 9:55:32 PM |  | Processor: 8 GenuineIntel        Intel(R) Core(TM) i7-2600K CPU @ 3.40GHz [Family 6 Model 42 Stepping 7]
    11/14/2012 9:55:32 PM |  | Processor: 256.00 KB cache
    11/14/2012 9:55:32 PM |  | 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 syscall nx lm vmx tm2 popcnt aes pbe
    11/14/2012 9:55:32 PM |  | OS: Microsoft Windows 8: Professional x64 Edition, (06.02.9200.00)
    11/14/2012 9:55:32 PM |  | Memory: 15.90 GB physical, 31.90 GB virtual
    11/14/2012 9:55:32 PM |  | Disk: 111.45 GB total, 12.49 GB free
    11/14/2012 9:55:32 PM |  | Local time is UTC -5 hours
    11/14/2012 9:55:32 PM |  | ATI GPU 0: Tahiti (CAL version 1.4.1741, 3072MB, 3032MB available, 9184 GFLOPS peak)
    11/14/2012 9:55:32 PM |  | ATI GPU 1: (not used) Capeverde (CAL version 1.4.1741, 1024MB, 984MB available, 3280 GFLOPS peak)
    11/14/2012 9:55:32 PM |  | OpenCL: ATI GPU 0: Tahiti (driver version 1084.2 (VM), device version OpenCL 1.2 AMD-APP (1084.2), 3072MB, 3032MB available)
    11/14/2012 9:55:32 PM |  | OpenCL: ATI GPU 1 (not used): Capeverde (driver version 1084.2 (VM), device version OpenCL 1.2 AMD-APP (1084.2), 1024MB, 984MB available)
    11/14/2012 9:55:32 PM |  | Config: don't compute while hl2.exe is running
    11/14/2012 9:55:32 PM | World Community Grid | URL http://www.worldcommunitygrid.org/; Computer ID 2204443; resource share 100
    11/14/2012 9:55:32 PM | World Community Grid | General prefs: from World Community Grid (last modified 10-Oct-2012 19:56:56)
    11/14/2012 9:55:32 PM | World Community Grid | Computer location: home
    11/14/2012 9:55:32 PM |  | General prefs: using separate prefs for home
    11/14/2012 9:55:32 PM |  | Reading preferences override file
    11/14/2012 9:55:32 PM |  | Preferences:
    11/14/2012 9:55:32 PM |  | max memory usage when active: 12208.41MB
    11/14/2012 9:55:32 PM |  | max memory usage when idle: 14650.09MB
    11/14/2012 9:55:32 PM |  | max disk usage: 10.00GB
    11/14/2012 9:55:32 PM |  | suspend work if non-BOINC CPU load exceeds 13 %
    11/14/2012 9:55:32 PM |  | (to change preferences, visit the web site of an attached project, or select Preferences in the Manager)
    11/14/2012 9:55:32 PM |  | Not using a proxy
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task SN2S_AAW88536_0000255_0626_0 using sn2s version 620 in slot 3
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task qj088_00106_0 using hpf2 version 640 in slot 2
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task faah36508_ZINC23136516_xh2_xtal_00_0 using faah version 640 in slot 5
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task HFCC_target-10_01316027_target-10_0001_0 using hfcc version 640 in slot 1
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task qj100_00011_4 using hpf2 version 640 in slot 6
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task HFCC_target-10_01316328_target-10_0000_1 using hfcc version 640 in slot 7
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task faah36513_ZINC02663138_xh2_xtal_00_0 using faah version 640 in slot 4
    11/14/2012 9:55:32 PM | World Community Grid | Restarting task X0930076700862200610260041_1 using hcc1 version 705 (ati_hcc1) in slot 0
    
     
    Last edited: Nov 15, 2012
    Crunching for Team TPU More than 25k PPD
  14. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    What's your exact cc_config file text? And I assume you hit "Read config file" after adding it? And where did you put it? It looks like BOINC isn't seeing the instructions to use all GPUs
     
    Jstn7477 says thanks.
    Crunching for Team TPU
  15. KieX

    Joined:
    Dec 18, 2009
    Messages:
    2,408 (1.36/day)
    Thanks Received:
    2,461
    Location:
    London, UK
    Can you copy the errors you get? If we coincide on FB or something can give you live help too.

    Can you post a copy of the cc_config.xml? If it's in the file it should appear in the log.
     
    Jstn7477 says thanks.
  16. Jstn7477

    Jstn7477

    Joined:
    Aug 30, 2009
    Messages:
    3,881 (2.06/day)
    Thanks Received:
    1,560
    Location:
    Sarasota, Florida, USA
    EDIT: I angrily uninstalled BOINC, nuked the BOINC folder, made an empty cc_config.xml with just the 4 sections and the one flag, and it looks like it is working!

    cc_config.xml in C:\ProgramData\BOINC

    Code:
    <cc_config>
        <log_flags>
            <file_xfer>1</file_xfer>
            <sched_ops>1</sched_ops>
            <task>1</task>
            <app_msg_receive>0</app_msg_receive>
            <app_msg_send>0</app_msg_send>
            <async_file_debug>0</async_file_debug>
            <benchmark_debug>0</benchmark_debug>
            <checkpoint_debug>0</checkpoint_debug>
            <coproc_debug>0</coproc_debug>
            <cpu_sched>0</cpu_sched>
            <cpu_sched_debug>0</cpu_sched_debug>
            <cpu_sched_status>0</cpu_sched_status>
            <dcf_debug>0</dcf_debug>
            <disk_usage_debug>0</disk_usage_debug>
            <priority_debug>0</priority_debug>
            <file_xfer_debug>0</file_xfer_debug>
            <gui_rpc_debug>0</gui_rpc_debug>
            <heartbeat_debug>0</heartbeat_debug>
            <http_debug>0</http_debug>
            <http_xfer_debug>0</http_xfer_debug>
            <mem_usage_debug>0</mem_usage_debug>
            <network_status_debug>0</network_status_debug>
            <poll_debug>0</poll_debug>
            <proxy_debug>0</proxy_debug>
            <rr_simulation>0</rr_simulation>
            <rrsim_detail>0</rrsim_detail>
            <sched_op_debug>0</sched_op_debug>
            <scrsave_debug>0</scrsave_debug>
            <slot_debug>0</slot_debug>
            <state_debug>0</state_debug>
            <statefile_debug>0</statefile_debug>
            <suspend_debug>0</suspend_debug>
            <task_debug>0</task_debug>
            <time_debug>0</time_debug>
            <trickle_debug>0</trickle_debug>
            <unparsed_xml>0</unparsed_xml>
            <work_fetch_debug>0</work_fetch_debug>
            <notice_debug>0</notice_debug>
        </log_flags>
        <options>
            <abort_jobs_on_exit>0</abort_jobs_on_exit>
            <allow_multiple_clients>0</allow_multiple_clients>
            <allow_remote_gui_rpc>0</allow_remote_gui_rpc>
            <client_version_check_url>http://boinc.berkeley.edu/download.php?xml=1</client_version_check_url>
            <client_download_url>http://boinc.berkeley.edu/download.php</client_download_url>
            <disallow_attach>0</disallow_attach>
            <dont_check_file_sizes>0</dont_check_file_sizes>
            <dont_contact_ref_site>0</dont_contact_ref_site>
            <exclusive_app>hl2.exe</exclusive_app>
            <exit_after_finish>0</exit_after_finish>
            <exit_before_start>0</exit_before_start>
            <exit_when_idle>0</exit_when_idle>
            <fetch_minimal_work>0</fetch_minimal_work>
            <force_auth>default</force_auth>
            <http_1_0>0</http_1_0>
            <http_transfer_timeout>300</http_transfer_timeout>
            <http_transfer_timeout_bps>10</http_transfer_timeout_bps>
            <max_file_xfers>8</max_file_xfers>
            <max_file_xfers_per_project>2</max_file_xfers_per_project>
            <max_stderr_file_size>0</max_stderr_file_size>
            <max_stdout_file_size>0</max_stdout_file_size>
            <max_tasks_reported>0</max_tasks_reported>
            <ncpus>-1</ncpus>
            <network_test_url>http://www.google.com/</network_test_url>
            <no_alt_platform>0</no_alt_platform>
            <no_gpus>0</no_gpus>
            <no_info_fetch>0</no_info_fetch>
            <no_priority_change>0</no_priority_change>
            <os_random_only>0</os_random_only>
    	<use_all_gpus>1</use_all_gpus>
    <proxy_info>
        <socks_server_name></socks_server_name>
        <socks_server_port>80</socks_server_port>
        <http_server_name></http_server_name>
        <http_server_port>80</http_server_port>
        <socks5_user_name></socks5_user_name>
        <socks5_user_passwd></socks5_user_passwd>
        <http_user_name></http_user_name>
        <http_user_passwd></http_user_passwd>
        <no_proxy></no_proxy>
    </proxy_info>
            <rec_half_life_days>10.000000</rec_half_life_days>
            <report_results_immediately>0</report_results_immediately>
            <run_apps_manually>0</run_apps_manually>
            <save_stats_days>30</save_stats_days>
            <skip_cpu_benchmarks>0</skip_cpu_benchmarks>
            <simple_gui_only>0</simple_gui_only>
            <start_delay>0</start_delay>
            <stderr_head>0</stderr_head>
            <suppress_net_info>0</suppress_net_info>
            <unsigned_apps_ok>0</unsigned_apps_ok>
            <use_all_gpus>0</use_all_gpus>
            <use_certs>0</use_certs>
            <use_certs_only>0</use_certs_only>
        </options>
    </cc_config>
    
     
    Last edited: Nov 15, 2012
    KieX says thanks.
    Crunching for Team TPU More than 25k PPD
  17. KieX

    Joined:
    Dec 18, 2009
    Messages:
    2,408 (1.36/day)
    Thanks Received:
    2,461
    Location:
    London, UK
    You have the code twice, once with 1 once with 0. Delete the one with 0 and you're sorted:
    (check code, highlighted red)

     
    Daimus and Jstn7477 say thanks.
  18. Jstn7477

    Jstn7477

    Joined:
    Aug 30, 2009
    Messages:
    3,881 (2.06/day)
    Thanks Received:
    1,560
    Location:
    Sarasota, Florida, USA
    Huh, can't believe I did a stupid mistake like that. Well, after all the trouble I have the latest beta BOINC version and 2 GPUs working. I appreciate the help KieX (and Ion, didn't see your ninja post until now)! :toast:
     
    [Ion] and KieX say thanks.
    Crunching for Team TPU More than 25k PPD
  19. KieX

    Joined:
    Dec 18, 2009
    Messages:
    2,408 (1.36/day)
    Thanks Received:
    2,461
    Location:
    London, UK
    No worries. :toast: Time to sit back and watch them roar now :rockout:
     
    Jstn7477 says thanks.
  20. brandonwh64

    brandonwh64 Addicted to Bacon and StarCrunches!!!

    Joined:
    Sep 6, 2009
    Messages:
    18,680 (9.97/day)
    Thanks Received:
    6,154
    Location:
    Chatsworth, GA
    After following o3m's and manofthem's guides and .XML files I have 6 WU per card on the 6950's!!!

    [​IMG]
     
    manofthem, Daimus and [Ion] say thanks.
    Crunching for Team TPU
  21. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    That's awesome! Keep us posted how long they end up taking--those preliminary results look incredible! :D
     
    Crunching for Team TPU
  22. t_ski

    t_ski Former Staff

    Joined:
    Apr 11, 2006
    Messages:
    10,909 (3.50/day)
    Thanks Received:
    4,390
    So are there any other projects that can utilize GPU WU's?
     
    Crunching for Team TPU
  23. Daimus

    Daimus

    Joined:
    Sep 20, 2011
    Messages:
    527 (0.47/day)
    Thanks Received:
    474
    Location:
    somewhere in Eastern Europe
    Unfortunately, HCC is so far the only one.
     
    t_ski says thanks.
  24. t_ski

    t_ski Former Staff

    Joined:
    Apr 11, 2006
    Messages:
    10,909 (3.50/day)
    Thanks Received:
    4,390
    Thanks - that's what I thought.
     
    Crunching for Team TPU
  25. [Ion]

    [Ion] WCG Team Assistant

    Joined:
    Sep 15, 2009
    Messages:
    11,855 (6.36/day)
    Thanks Received:
    11,082
    Location:
    North Carolina, United States
    Well, I had my first issue where the GPU WUs just get "stuck" and don't do anything. I lost about an hour while I had two just sit there and spin...I aborted them. Any idea how to avoid this in the future?
     
    Crunching for Team TPU

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page