| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: inefficient cpu? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 11
|
|
| Author |
|
|
Paneologe
Cruncher Joined: Sep 16, 2010 Post Count: 11 Status: Offline Project Badges: 
|
Hallo,
----------------------------------------I installed some "new" PCs, running this project, i get very little points in comparison to claimed points: Result Name Device Name Status Sent Time Time Due / Return Time CPU Time / Elapsed Time (hours) Claimed/ Granted BOINC Credit E219766_ 352_ K.21.C15FH8N3S2.00226834.2.set1d06_ 1-- epsilon Valid 3/8/14 05:09:34 3/11/14 06:04:11 4.62 / 7.53 165.5 / 50.1 E219726_ 356_ K.21.C15FH8N3O2.00336008.3.set1d06_ 2-- delta Valid 3/6/14 23:16:42 3/10/14 23:10:48 4.61 / 7.82 164.9 / 20.1 E219725_ 093_ K.19.C16FH9SSe.00371853.2.set1d06_ 2-- delta Valid 3/6/14 21:43:49 3/10/14 16:45:47 4.93 / 9.19 194.5 / 36.0 E219716_ 285_ K.21.C15FH6NO2S2.00293911.set1d06_ 0-- delta Valid 3/6/14 15:10:54 3/10/14 04:24:39 4.42 / 6.73 139.7 / 47.4 cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz stepping : 2 microcode : 0x5d no similar problems on other machines using AMD 64bit or VIA C7 processors or with mentioned cpus on other projects I am wondering wheter it would be better to exclude those from this project, if they do better on other projects? any hints? thanks [Edit 1 times, last edit by Paneologe at Mar 11, 2014 6:30:16 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Look at your CPU time Vs your elapsed time, elapsed is nearly twice CPU, very inefficient. That is why your points are low I think.
 |
||
|
|
jonnieb-uk
Ace Cruncher England Joined: Nov 30, 2011 Post Count: 6105 Status: Offline Project Badges:
|
Look at your CPU time Vs your elapsed time, elapsed is nearly twice CPU, very inefficient. That is why your points are low I think. If you browse the CEP2 forum you'll see that Points for CEP2 work has been a topic for at least the last 3 months. The problem is not especially your machines - Points claimed were about normal. i get very little points in comparison to claimed points Ir's 'luck of the draw' what wingman you get matched with.It would be interesting to see the Wingman's result alongside your results. The problem will disappear once you have submitted enough valid results (for each machine) to qualify for zero quorum work = no wingman. How many CEP2 WUs are you running concurrently on each machine? CEP2 makes high demands on the CPU and running multiple WUs can degrade peformance. It would also be useful to know what BOINC Vers and OS you have on the machines. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello Paneologe,
Look at your CPU time Vs your elapsed time, elapsed is nearly twice CPU, very inefficient. That is why your points are low I think. Scribe has spotted the main problem, I think. It is almost as though you are trying to run CEP2 on virtual cores instead of dedicating a real core to them. CEP2 has a lot of I/O inefficiency in programs compiled to run on Windows, so you have to minimize the inefficiency. If you posted the start of your Event Log, which describes your system, people might be able to advise you. Particularly if you described what you are running on your system. Lawrence |
||
|
|
Paneologe
Cruncher Joined: Sep 16, 2010 Post Count: 11 Status: Offline Project Badges:
|
Thanks so far.
----------------------------------------Additional info about these (identical) systems: delta:~ # cat /proc/version Linux version 3.11.10-7-pae (geeko@buildhost) (gcc version 4.8.1 20130909 [gcc-4_8-branch revision 202388] (SUSE Linux) ) #1 SMP Mon Feb 3 09:41:24 UTC 2014 (750023e) delta:~ # rpm -q boinc-client boinc-client-7.2.41-1.1.i586 delta:~ # cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz stepping : 2 microcode : 0x5d cpu MHz : 1867.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : 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 ht tm pbe nx lm constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm lahf_lm dtherm tpr_shadow bogomips : 3724.21 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management: processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz stepping : 2 microcode : 0x5d cpu MHz : 1867.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 apicid : 1 initial apicid : 1 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : 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 ht tm pbe nx lm constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm lahf_lm dtherm tpr_shadow bogomips : 3724.21 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management: delta:~ # boinccmd --get_state ======== Tasks ======== 1) ----------- name: MCM1_0002797_2735_1 WU name: MCM1_0002797_2735 project URL: http://www.worldcommunitygrid.org/ report deadline: Sun Mar 16 08:09:30 2014 ready to report: no got server ack: no final CPU time: 0.000000 state: downloaded scheduler state: scheduled exit_status: 0 signal: 0 suspended via GUI: no active_task_state: SUSPENDED app version num: 728 checkpoint CPU time: 22833.750000 current CPU time: 23290.410000 fraction done: 0.561527 swap size: 0.000000 working set size: 0.000000 estimated CPU time remaining: 7236.941190 2) ----------- name: E219815_926_K.21.C19FH11O.00308644.1.set1d06_0 WU name: E219815_926_K.21.C19FH11O.00308644.1.set1d06 project URL: http://www.worldcommunitygrid.org/ report deadline: Wed Mar 19 20:53:20 2014 ready to report: no got server ack: no final CPU time: 0.000000 state: downloaded scheduler state: scheduled exit_status: 0 signal: 0 suspended via GUI: no active_task_state: SUSPENDED app version num: 640 checkpoint CPU time: 6324.800000 current CPU time: 6494.576000 fraction done: 0.100225 swap size: 0.000000 working set size: 0.000000 estimated CPU time remaining: 11334.174684 /var/log/boinc_client.log 08-Mar-2014 15:30:52 [---] Starting BOINC client version 7.2.41 for i686-pc-linux-gnu 08-Mar-2014 15:30:52 [---] log flags: file_xfer, sched_ops, task 08-Mar-2014 15:30:52 [---] Libraries: libcurl/7.32.0 OpenSSL/1.0.1e zlib/1.2.8 libidn/1.25 libssh2/1.4.3 08-Mar-2014 15:30:52 [---] Data directory: /var/lib/boinc 08-Mar-2014 15:30:52 [---] No usable GPUs found 08-Mar-2014 15:30:52 [---] Host name: delta 08-Mar-2014 15:30:52 [---] Processor: 2 GenuineIntel Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz [Family 6 Model 15 Stepping 2] 08-Mar-2014 15:30:52 [---] 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 ht tm pbe nx lm constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm lahf_lm dtherm tpr_shadow 08-Mar-2014 15:30:52 [---] OS: Linux: 3.11.10-7-pae 08-Mar-2014 15:30:52 [---] Memory: 1.96 GB physical, 16.01 GB virtual 08-Mar-2014 15:30:52 [---] Disk: 167.33 GB total, 132.30 GB free 08-Mar-2014 15:30:52 [---] Local time is UTC +1 hours 08-Mar-2014 15:30:52 [---] Config: GUI RPCs allowed from: 08-Mar-2014 15:30:52 [---] alpha.swabian.net 08-Mar-2014 15:30:52 [World Community Grid] URL http://www.worldcommunitygrid.org/; Computer ID 2845745; resource share 150 08-Mar-2014 15:30:52 [World Community Grid] General prefs: from World Community Grid (last modified 01-Mar-2014 07:35:43) 08-Mar-2014 15:30:52 [World Community Grid] Host location: none 08-Mar-2014 15:30:52 [World Community Grid] General prefs: using your defaults 08-Mar-2014 15:30:52 [---] Reading preferences override file 08-Mar-2014 15:30:52 [---] Preferences: 08-Mar-2014 15:30:52 [---] max memory usage when active: 1002.64MB 08-Mar-2014 15:30:52 [---] max memory usage when idle: 1804.75MB 08-Mar-2014 15:30:52 [---] max disk usage: 83.67GB 08-Mar-2014 15:30:52 [---] (to change preferences, visit a project web site or select Preferences in the Manager) 08-Mar-2014 15:30:52 [---] Using proxy info from GUI 08-Mar-2014 15:30:52 [---] Using HTTP proxy proxy.swabian.net:3128 08-Mar-2014 15:30:54 Initialization completed maximum 2 jobs are running concurrently, one for each cpu core. nothing else (costly relevant) running on these machines - not even X. machines are diskless with root on nfs, this causes sometimes high I/O wait, that's why cpu time is much lower than the elapsed time. But for the points only the cpu time should be used. something like cputime*cpubenchmark... I have very similar installs using VIA C7 processors, also with nfsroot and high I/O wait, elapsed time is twice cpu time, but claiming points "ordinary".. Edit: cpu benchmarks: 10-Mar-2014 19:23:19 [---] Running CPU benchmarks 10-Mar-2014 19:23:19 [---] Suspending computation - CPU benchmarks in progress 10-Mar-2014 19:23:52 [---] Benchmark results: 10-Mar-2014 19:23:52 [---] Number of CPUs: 2 10-Mar-2014 19:23:52 [---] 2099 floating point MIPS (Whetstone) per CPU 10-Mar-2014 19:23:52 [---] 5873 integer MIPS (Dhrystone) per CPU 10-Mar-2014 19:23:54 [---] Resuming computation [Edit 1 times, last edit by Paneologe at Mar 12, 2014 7:12:12 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Ouch. Windows is notoriously inefficient on CEP2 I/O but Linux does much better. Does anybody have any ideas?
Lawrence |
||
|
|
Paneologe
Cruncher Joined: Sep 16, 2010 Post Count: 11 Status: Offline Project Badges:
|
just switch to linux
|
||
|
|
Paneologe
Cruncher Joined: Sep 16, 2010 Post Count: 11 Status: Offline Project Badges:
|
processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz stepping : 2 microcode : 0x5d cpu MHz : 1867.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : 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 ht tm pbe nx lm constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm lahf_lm dtherm tpr_shadow bogomips : 3724.21 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management: just noticed that's a 64bit cpu... and i installed 32bit os... going to install 64bit... |
|||
|
||||
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7668 Status: Offline Project Badges: 
|
08-Mar-2014 15:30:52 [---] max memory usage when active: 1002.64MB Since CEP2 is notoriously I/O bound, I think you are causing a lot of thrashing into and out of virtual memory because you are only allowing 1 gb of actual memory for having two tasks running. You have plenty of memory in this machine, allow more actual memory to be used when the machine is active. Increase this setting. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Paneologe
Cruncher Joined: Sep 16, 2010 Post Count: 11 Status: Offline Project Badges:
|
the problem completeley disappeared after installing an 64bit os :)
applying the settings for memory usage (max memory usage when active/idle: 1797.73MB) didn't influence the granted points. |
||
|
|
|