1. Este site usa cookies. Ao continuar a usar este site está a concordar com o nosso uso de cookies. Saber Mais.
  1. transistor

    transistor I fold therefore I AM

    Ora viva,

    Bem, há umas semanas resolvi utilizar o cliente SMP windows no pc de casa, que na altura corria 3 consolas de cliente normal e uma consola de cliente gráfico.

    Configurei a consola SMP, tudo ok, mas para meu espanto noutro dia ao olhar para os meus stats, descobri que já contava com mais de 2000 WU realizadas :wow:


    http://folding.extremeoverclocking.com/user_summary.php?s=&u=240642


    Lá fui ver o que se passava, e vi que o cliente gráfico recebe as wu's mas depois apenas processa 1% e envia o resultado para os servers.

    Deixo aqui um log:

    Código:
    # Windows Console GPU Edition #################################################
    ###############################################################################
    
                           [email protected] Client Version 6.11beta5
    
                              http://folding.stanford.edu
    
    ###############################################################################
    ###############################################################################
    
    Launch directory: C:\Users\Rui\AppData\Roaming\[email protected]
    
    
    [19:40:19] - Ask before connecting: No
    [19:40:19] - User name: transistor_pt (Team 35271)
    [19:40:19] - User ID: 3A75BE93155DBC3E
    [19:40:19] - Machine ID: 2
    [19:40:19] 
    [19:40:19] Loaded queue successfully.
    [19:40:19] Initialization complete
    [19:40:19] + Benchmarking ...
    [19:40:19] 
    [19:40:19] + Processing work unit
    [19:40:19] Core required: FahCore_11.exe
    [19:40:19] Core found.
    [19:40:19] Working on Unit 09 [June 1 19:40:19]
    [19:40:19] + Working ...
    [19:40:20] 
    [19:40:20] *------------------------------*
    [19:40:20] [email protected] GPU Core - Beta
    [19:40:20] Version 1.02 (Fri May 30 15:56:39 PDT 2008)
    [19:40:20] 
    [19:40:20] Compiler  : 
    [19:40:20] Build host: amoeba
    [19:40:20] Preparing to commence simulation
    [19:40:20] - Looking at optimizations...
    [19:40:20] - Files status OK
    [19:40:20] - Expanded 43454 -> 236913 (decompressed 545.2 percent)
    [19:40:20] Called DecompressByteArray: compressed_data_size=43454 data_size=236913, decompressed_data_size=236913 diff=0
    [19:40:20] - Digital signature verified
    [19:40:20] 
    [19:40:20] Project: 4709 (Run 1, Clone 55, Gen 27)
    [19:40:20] 
    [19:40:20] Assembly optimizations on if available.
    [19:40:20] Entering M.D.
    [19:40:26] Will resume from checkpoint file
    [19:40:26] Working on 544 p4709_fip35_ww_domain
    [19:40:26] Client config found, loading data.
    [19:40:26] Starting GUI Server
    [19:40:32] Resuming from checkpoint
    [19:40:32] Verified work/wudata_09.log
    [19:40:32] Verified work/wudata_09.edr
    [19:40:32] Verified work/wudata_09.trr
    [19:40:32] Verified work/wudata_09.xtc
    [19:40:32] mdrun_gpu returned -1
    [19:40:32] Going to send back what have done.
    [19:40:33] logfile size: 13707 info=13707 bed=25 hdr=1
    [19:40:33] - Writing 14245 bytes of core data to disk...
    [19:40:33] Done: 13733 -> 4688 (compressed to 34.1 percent)
    [19:40:33]   ... Done.
    [19:40:33] 
    [19:40:33] [email protected] Core Shutdown: EARLY_UNIT_END
    [19:40:36] CoreStatus = 72 (114)
    [19:40:36] Sending work to server
    [19:40:36] - Read packet limit of 540015616... Set to 524286976.
    
    
    [19:40:36] + Attempting to send results
    [19:40:37] + Results successfully sent
    [19:40:37] Thank you for your contribution to [email protected]
    [19:40:41] - Preparing to get new work unit...
    [19:40:41] + Attempting to get work packet
    [19:40:41] - Connecting to assignment server
    [19:40:42] - Successful: assigned to (171.64.65.103).
    [19:40:42] + News From [email protected]: GPU folding beta
    [19:40:42] Loaded queue successfully.
    [19:40:44] + Closed connections
    [19:40:49] 
    [19:40:49] + Processing work unit
    [19:40:49] Core required: FahCore_11.exe
    [19:40:49] Core found.
    [19:40:49] Working on Unit 00 [June 1 19:40:49]
    [19:40:49] + Working ...
    [19:40:49] 
    [19:40:49] *------------------------------*
    [19:40:49] [email protected] GPU Core - Beta
    [19:40:49] Version 1.02 (Fri May 30 15:56:39 PDT 2008)
    [19:40:49] 
    [19:40:49] Compiler  : 
    [19:40:49] Build host: amoeba
    [19:40:49] Preparing to commence simulation
    [19:40:49] - Looking at optimizations...
    [19:40:49] - Created dyn
    [19:40:49] - Files status OK
    [19:40:49] - Expanded 43385 -> 236913 (decompressed 546.0 percent)
    [19:40:49] Called DecompressByteArray: compressed_data_size=43385 data_size=236913, decompressed_data_size=236913 diff=0
    [19:40:49] - Digital signature verified
    [19:40:49] 
    [19:40:49] Project: 4709 (Run 0, Clone 41, Gen 43)
    [19:40:49] 
    [19:40:49] Assembly optimizations on if available.
    [19:40:49] Entering M.D.
    [19:40:55] Working on 544 p4709_fip35_ww_domain
    [19:40:56] Client config found, loading data.
    [19:40:56] Starting GUI Server
    [19:43:14] Completed 1%
    [19:43:14] mdrun_gpu returned -1
    [19:43:14] Going to send back what have done.
    [19:43:15] logfile size: 14215 info=14215 bed=25 hdr=1
    [19:43:15] - Writing 14753 bytes of core data to disk...
    [19:43:15] Done: 14241 -> 4731 (compressed to 33.2 percent)
    [19:43:15]   ... Done.
    [19:43:15] 
    [19:43:15] [email protected] Core Shutdown: EARLY_UNIT_END
    [19:43:18] CoreStatus = 72 (114)
    [19:43:18] Sending work to server
    [19:43:18] - Read packet limit of 540015616... Set to 524286976.
    
    
    [19:43:18] + Attempting to send results
    [19:43:19] + Results successfully sent
    [19:43:19] Thank you for your contribution to [email protected]
    [19:43:23] - Preparing to get new work unit...
    [19:43:23] + Attempting to get work packet
    [19:43:23] - Connecting to assignment server
    [19:43:24] - Successful: assigned to (171.64.65.103).
    [19:43:24] + News From [email protected]: GPU folding beta
    [19:43:24] Loaded queue successfully.
    [19:43:26] + Closed connections
    [19:43:31] 
    [19:43:31] + Processing work unit
    [19:43:31] Core required: FahCore_11.exe
    [19:43:31] Core found.
    [19:43:31] Working on Unit 01 [June 1 19:43:31]
    [19:43:31] + Working ...
    [19:43:31] 
    [19:43:31] *------------------------------*
    [19:43:31] [email protected] GPU Core - Beta
    [19:43:31] Version 1.02 (Fri May 30 15:56:39 PDT 2008)
    [19:43:31] 
    [19:43:31] Compiler  : 
    [19:43:31] Build host: amoeba
    [19:43:31] Preparing to commence simulation
    [19:43:31] - Looking at optimizations...
    [19:43:31] - Created dyn
    [19:43:31] - Files status OK
    [19:43:31] - Expanded 43450 -> 236913 (decompressed 545.2 percent)
    [19:43:31] Called DecompressByteArray: compressed_data_size=43450 data_size=236913, decompressed_data_size=236913 diff=0
    [19:43:31] - Digital signature verified
    [19:43:31] 
    [19:43:31] Project: 4709 (Run 2, Clone 6, Gen 48)
    [19:43:31] 
    [19:43:31] Assembly optimizations on if available.
    [19:43:31] Entering M.D.
    [19:43:37] Working on 544 p4709_fip35_ww_domain
    [19:43:38] Client config found, loading data.
    [19:43:38] Starting GUI Server
    Já tentei correr apenas a consola GPU, sem a smp ligada, e o problema mantém-se. Já tentei desistalar o re-instalar a consola GPU e nada...

    Alguém é capaz de me dar aqui uma ajudinha neste problema?


    Já agora, algém me sabe dizer se é possivel correr 2 consolas gpu na mesma máquina, ie, utilizar 2 placas gráficas distintas cada qual com 1 cliente, ou se o Xfire é aproveitado?


    Um abraço e :kfold:
    Rui
     
    Última edição: 4 de Junho de 2008
  2. BadPlayer

    BadPlayer I folded Myself

    Qual é a versão do CCC que estás a usar?
    Se for a mais recente passa para a 8.4, pois a 8.5 dá muitos EUEs.
    Mesmo com 2 placas gráficas ainda não são suportados 2 clientes GPUs a funcionar em simultâneo.

    :kfold:
     
  3. transistor

    transistor I fold therefore I AM

    Usava os 8.4 quando se deu isto pela primeira vez....
    Por acaso agora tenho os 8.5, instalei de propósito para confirmar se seria do driver, logo já vejo isso, já faço o rollback do driver...

    Algum outro conselho\consideração para ver se consigo por isto a trabalhar em condições novamente? Os meus ppd's sem a consola gpu andam miseráveis :lol:

    Um abraço
     
  4. transistor

    transistor I fold therefore I AM

    Bem, voltei ontem ás cat 8.4, e isto continua... :mad:

    Alguém é capaz de me dar uma ajudinha nisto?

    Um abraço
     
  5. transistor

    transistor I fold therefore I AM

    Bem, lá consegui resolver isto...

    O problema dos loops sucessivos de envio de WU's a 1% foi resolvido com uma limpeza do Registo do Windows e re-instalação do cliente referente ao cliente gráfico claro está...

    So far so good...

    Aproveito para dizer que as FireGL baseadas no R600 e RVXXX trabalham também com o cliente, e por sinal muito bem (pelo menos as V8600 :)).

    Um abraço e :kfold:
     

Partilhar esta Página