World Community Grid - View Thread

World Community Grid Forums

Category: Active Research

Forum: Mapping Cancer Markers Forum

Thread: CPU Time 0.00

Quick Go »

No member browsing this thread

Thread Status: Active
Total posts in this thread: 19

[ ]

Author

This topic has been viewed 6431 times and has 18 replies

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


CPU Time 0.00

I am currently receiving a CPU Time of 0.00 for all MCM WUs returned. It appears the WUs are valid, just no CPU time is being reported.

Running WU / Completed WU

Has anyone had this issue before?

Currently running Gentoo Linux x64 with a custom compiled version of Boinc client 7.2.20

[Feb 27, 2014 10:52:12 PM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

Hi Jagaer,
I can't even run a task without getting more elapsed time than CPU time. Somebody else will have to solve this mystery.

confused

Lawrence

[Feb 27, 2014 11:43:39 PM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

Could it be related to the custom compiled version of the boinc client?

[Feb 28, 2014 2:30:42 AM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

It could be related to the custom version, but if you look at the running task there is more CPU time than elapsed time, so I'm wondering if the server side is rejecting as impossible.

I had the same issue with CEP2 (more CPU time than elapsed time)

The only thing I can think of is that it doesn't know what to do with the MCM found in G34 (2xInt Cores with 1xFP Core) or the All Core Turbo is causing a time skew.

As long as the WUs are returning correctly, it isn't a big deal, but thought someone may have encountered this issue before.

I'll try compiling a newer version. If that doesn't work, I'll load a stock binary. (Trying to avoid a non-optimized binary as the AMD MCM doesn't run nearly as efficiently without GCC BDVER)

----------------------------------------
[Edit 1 times, last edit by Former Member at Feb 28, 2014 2:48:47 AM]

[Feb 28, 2014 2:48:20 AM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

Thanks Jagaer,

The only thing I can think of is that it doesn't know what to do with the MCM found in G34 (2xInt Cores with 1xFP Core) or the All Core Turbo is causing a time skew.

I think that this explains the extremely puzzling CPU/Elapsed paradoxes I have recently been coming across.

biggrin

Lawrence

[Feb 28, 2014 2:38:59 PM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

I upgraded to a custom compiled version of 7.2.41 x86_64-pc-linux-gnu, but it is still exhibiting the same issue.

I perform an installation and reboot prior to starting the software and let it run for at least 24 hours before checking the CPU Time.

I'll install the latest precompiled version from BOINC. If that doesn't fix it, it must be something to do with my distro (CPU Flag, "time" command etc...)

[Mar 3, 2014 12:52:10 AM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

Hi Jagaer,
I doubt that this is a BOINC issue as such, at least npt a coding bug. My new hypothesis is that the new cores with 2 integer units and 1 FPU behave differently in the way they report time. So the actual algorithm for handling time needs changing, assuming that the data reported from the new cores allows us to figure out the time used the way that we are accustomed to. This will take some research by the BOINC Development Team. I will drop a link to this thread to our staff, for them to decide if I am having a useless brainstorm or not.

Lawrence

[Mar 3, 2014 3:16:30 AM]

armstrdj
Former World Community Grid Tech
Joined: Oct 21, 2004
Post Count: 695
Status: Offline
Project Badges:

5 year badge for Human Proteome Folding - Phase 2

14 day badge for Help Cure Muscular Dystrophy

90 day badge for Discovering Dengue Drugs - Together

90 day badge for Nutritious Rice for the World

90 day badge for The Clean Energy Project

2 year badge for Help Fight Childhood Cancer

90 day badge for Influenza Antiviral Drug Search

2 year badge for Help Cure Muscular Dystrophy - Phase 2

2 year badge for Discovering Dengue Drugs - Together - Phase 2

2 year badge for The Clean Energy Project - Phase 2

2 year badge for Computing for Clean Water

2 year badge for Drug Search for Leishmaniasis

2 year badge for GO Fight Against Malaria

2 year badge for Computing for Sustainable Water

10 year badge for Mapping Cancer Markers

2 year badge for Uncovering Genome Mysteries

2 year badge for Outsmart Ebola Together

2 year badge for FightAIDS@Home - Phase 2

2 year badge for Microbiome Immunity Project

2 year badge for Africa Rainfall Project

2 year badge for OpenPandemics - COVID-19


Re: CPU Time 0.00

Let me know how the precomipled version runs. I need to go through the portion of the BOINC code related to CPU time for another issue and will let you know if I see anything.

Thanks,
armstrdj

[Mar 3, 2014 5:11:29 PM]

Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline


Re: CPU Time 0.00

Changed to precompiled boinc 7.2.42 downloaded from the BOINC page.

It appears that FAAH-Vina and CEP2 have CPU Time in advance of Elapsed time and FAAH-Autodock and MCM are not able to track CPU Time.

03-Mar-2014 17:08:29 [---] cc_config.xml not found - using defaults
03-Mar-2014 17:08:29 [---] Starting BOINC client version 7.2.42 for x86_64-pc-linux-gnu
03-Mar-2014 17:08:29 [---] log flags: file_xfer, sched_ops, task
03-Mar-2014 17:08:29 [---] Libraries: libcurl/7.35.0 OpenSSL/1.0.1f zlib/1.2.8
03-Mar-2014 17:08:29 [---] Data directory: /home/james/boincram
boinc: /usr/lib64/libssl.so.1.0.0: no version information available (required by boinc)
boinc: /usr/lib64/libcurl.so.4: no version information available (required by boinc)
boinc: /usr/lib64/libcrypto.so.1.0.0: no version information available (required by boinc)
03-Mar-2014 17:08:29 [---] CUDA: NVIDIA GPU 0: GeForce GTX 580 (driver version unknown, CUDA version 6.0, compute capability 2.0, 3072MB, 2916MB available, 1581 GFLOPS peak)
03-Mar-2014 17:08:29 [---] OpenCL: NVIDIA GPU 0: GeForce GTX 580 (driver version 331.20, device version OpenCL 1.1 CUDA, 3072MB, 2916MB available, 1581 GFLOPS peak)
03-Mar-2014 17:08:29 [---] Host name: jagaer
03-Mar-2014 17:08:29 [---] Processor: 32 AuthenticAMD AMD Eng Sample, ZS222445TGG44_32/22/2_2/16 [Family 21 Model 1 Stepping 0]
03-Mar-2014 17:08:29 [---] Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc extd_apicid amd_dcm aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 nodeid_msr topoext perfctr_core perfctr_nb arat cpb hw_pstate npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
03-Mar-2014 17:08:29 [---] OS: Linux: 3.11.4-pf
03-Mar-2014 17:08:29 [---] Memory: 62.91 GB physical, 0 bytes virtual
03-Mar-2014 17:08:29 [---] Disk: 10.00 GB total, 8.40 GB free
03-Mar-2014 17:08:29 [---] Local time is UTC -5 hours
03-Mar-2014 17:08:29 [---] VirtualBox version: 4.3.8_Gentoor92456
03-Mar-2014 17:08:29 [---] Version change (7.2.41 -> 7.2.42)
03-Mar-2014 17:08:29 [World Community Grid] URL http://www.worldcommunitygrid.org/; Computer ID 2787839; resource share 100
03-Mar-2014 17:08:29 [World Community Grid] General prefs: from World Community Grid (last modified 26-Feb-2014 06:13:26)
03-Mar-2014 17:08:29 [World Community Grid] Host location: none
03-Mar-2014 17:08:29 [World Community Grid] General prefs: using your defaults
03-Mar-2014 17:08:29 [---] Reading preferences override file
03-Mar-2014 17:08:29 [---] Preferences:
03-Mar-2014 17:08:29 [---] max memory usage when active: 48312.60MB
03-Mar-2014 17:08:29 [---] max memory usage when idle: 57975.12MB
03-Mar-2014 17:08:29 [---] max disk usage: 9.50GB
03-Mar-2014 17:08:29 [---] don't use GPU while active
03-Mar-2014 17:08:29 [---] (to change preferences, visit a project web site or select Preferences in the Manager)
03-Mar-2014 17:08:29 [---] Not using a proxy
03-Mar-2014 17:08:29 Initialization completed
03-Mar-2014 17:08:29 [---] Running CPU benchmarks
03-Mar-2014 17:08:29 [---] Suspending computation - CPU benchmarks in progress
03-Mar-2014 17:09:00 [---] Benchmark results:
03-Mar-2014 17:09:00 [---] Number of CPUs: 32
03-Mar-2014 17:09:00 [---] 1441 floating point MIPS (Whetstone) per CPU
03-Mar-2014 17:09:00 [---] 6839 integer MIPS (Dhrystone) per CPU
03-Mar-2014 17:09:01 [---] Suspending GPU computation - computer is in use
03-Mar-2014 17:17:17 [World Community Grid] Computation for task MCM1_0002646_3023_1 finished
03-Mar-2014 17:17:17 [World Community Grid] Starting task FAHV_x3ZT1_A_IN_FBPa_rig_0201374_0705_2

System Info:
/proc/version

cat /proc/version
Linux version 3.11.4-pf (root@jagaer) (gcc version 4.7.3 (Gentoo 4.7.3-r1 p1.3, pie-0.5.5) ) #1 SMP PREEMPT Mon Nov 4 09:42:15 EST 2013

cat /etc/make.conf
# These settings were set by the catalyst build script that automatically built this stage
CFLAGS="-O2 -pipe -march=native"
CHOST="x86_64-pc-linux-gnu"
USE="amd64 kde qt4 qt3support alsa phonon oss multilib cdr dvd dvdr dvdread xine \
audacious usb udev dbus pam opengl xv mp3 ogg vorbis libsamplerate mad quicktime \
real wmp xvid x264 ffmpeg win32codecs nptl nptlonly lame encode dts a52 aac \
subversion X truetype history cups foomatic python mysql sqlite extensions sql \
nsplugin webkit mng accessibility declarative zip samba bzip2 jpeg gif tiff svg \
png lm_sensors caps gstreamer extras consolekit policykit tools \
virtualbox evdev vdpau vaapi ntfsprogs \
-dso -sis -rage128 -radeon -matrox -3dfx -gamma -i8x0 -gnome -gtk -hal -arts -pulseaudio -semantic-desktop"
VIDEO_CARDS="nvidia"
CXXFLAGS="${CFLAGS}"
MAKEOPTS="-j32"

CPU Info

TurionPowerControl -l
TurionPowerControl 0.44-rc2 (trunk-r177M)
Turion Power States Optimization and Control - by blackshard

Main processor is Family 15h (Bulldozer/Interlagos/Valencia) Processor
Family: 0xf Model: 0x1 Stepping: 0x0
Extended Family: 0x15 Extended Model: 0x1
Package Type: 0x3 BrandId: 0x0
Machine has 4 nodes
Processor has 8 cores
Processor has 7 p-states
Processor has 1 boost states

Power States table:
-- Node: 0 Core 0
core 0 pstate 0 (pb0) - Boost PState Disabled
core 0 pstate 1 (p0) - En:1 VID:15 FID:22 DID:0.00 NBPS:0 Freq:3800 VCore:1.3625
core 0 pstate 2 (p1) - En:1 VID:43 FID:6 DID:0.00 NBPS:0 Freq:2200 VCore:1.0125
core 0 pstate 3 (p2) - En:1 VID:45 FID:4 DID:0.00 NBPS:1 Freq:2000 VCore:0.9875
core 0 pstate 4 (p3) - En:1 VID:47 FID:2 DID:0.00 NBPS:0 Freq:1800 VCore:0.9625
core 0 pstate 5 (p4) - En:1 VID:48 FID:0 DID:0.00 NBPS:0 Freq:1600 VCore:0.9500
core 0 pstate 6 (p5) - En:1 VID:50 FID:12 DID:1.00 NBPS:0 Freq:1400 VCore:0.9250

Memory Info:

TurionPowerControl -dram
TurionPowerControl 0.44-rc2 (trunk-r177M)
Turion Power States Optimization and Control - by blackshard

DRAM Configuration Status

Node 0 ---
DCT0: DDR3 frequency: 1600 MHz
Tcl=11 Trcd=11 Trp=11 Tras=28 Access Mode:1T Trtp=6 Trc=39 Twr=12 Trrd=5 Tcwl=8 Tfaw=24
TrwtWB=10 TrwtTO=9 Twtr=6 Twrrd=1 Twrwrsdsc=1 Trdrdsdsc=1 Tref=2 Trfc0=4 Trfc1=3 Trfc2=4 Trfc3=4
MaxRdLatency0=65 DataTxFifoWrDly0=0 RdPtrInit0=6
MaxRdLatency1=65 DataTxFifoWrDly1=0 RdPtrInit1=6
LDIMM0=EMPTY/EMPTY LDIMM1=OK/OK LDIMM2=EMPTY/EMPTY LDIMM3=EMPTY/EMPTY

DCT1: DDR3 frequency: 1600 MHz
Tcl=11 Trcd=11 Trp=11 Tras=28 Access Mode:1T Trtp=6 Trc=39 Twr=12 Trrd=5 Tcwl=8 Tfaw=24
TrwtWB=10 TrwtTO=9 Twtr=6 Twrrd=1 Twrwrsdsc=1 Trdrdsdsc=1 Tref=2 Trfc0=4 Trfc1=3 Trfc2=4 Trfc3=4
MaxRdLatency0=64 DataTxFifoWrDly0=0 RdPtrInit0=6
MaxRdLatency1=64 DataTxFifoWrDly1=0 RdPtrInit1=6
LDIMM0=EMPTY/EMPTY LDIMM1=OK/OK LDIMM2=EMPTY/EMPTY LDIMM3=EMPTY/EMPTY

Kernel Config: http://www.snapshotoflife.com/wcg/config-3.11.4-pf

----------------------------------------
[Edit 1 times, last edit by Former Member at Mar 3, 2014 11:26:07 PM]

[Mar 3, 2014 11:22:02 PM]

armstrdj
Former World Community Grid Tech
Joined: Oct 21, 2004
Post Count: 695
Status: Offline
Project Badges:


Re: CPU Time 0.00

FAAH-VINA and CEP2 are compound apps and therefore the CPU time is calculated differently. Those projects fork a sub process and the main application uses the proc filesystem to get the cpu time. The other projects don't do this and just use getrusage to get the cpu time. You can see this code in the BOINC source code under boinc_api.cpp in the boinc_worker_thread_cpu_time function. Do you run any other BOINC projects on this machine outside of World Community Grid? If so do they report the cpu time accurately?

Thanks,
armstrdj

[Mar 5, 2014 12:00:32 AM]

[ ]