Validation Inconclusive - Mac CUDA90 App - 38 V.I.!!!

Message boards : Number crunching : Validation Inconclusive - Mac CUDA90 App - 38 V.I.!!!
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969047 - Posted: 7 Dec 2018, 3:20:58 UTC
Last modified: 7 Dec 2018, 3:23:15 UTC

This is the MOST V.I.s that I've EVER seen on an NVIDIA Card!

MOST of these hit today! What happened??? Do I need to worry about this? Until today, my V.I.s were hanging around 7-10...


TL

[EDIT:]

762 into 38 = 4.9+ %

I've been normally hanging around 1.3 to 1.8 % until today.
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969047 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 17783
Credit: 406,467,362
RAC: 155,221
United Kingdom
Message 1969059 - Posted: 7 Dec 2018, 5:48:55 UTC

If I were you I'd be more worried by the pile of INVALID tasks that "appeared" on the 5th....
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1969059 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969096 - Posted: 7 Dec 2018, 15:55:26 UTC - in response to Message 1969059.  

If I were you I'd be more worried by the pile of INVALID tasks that "appeared" on the 5th....

Thanks, didn't see those! OMG!!! WTF???

Here's a StdErr Report from one of the Invalids:

Task 7209724488

Name 04dc18aa.12315.20517.11.38.140.vlar_1
Workunit 3248950433
Created 5 Dec 2018, 5:02:38 UTC
Sent 5 Dec 2018, 5:28:35 UTC
Report deadline 27 Jan 2019, 10:28:17 UTC
Received 5 Dec 2018, 12:47:28 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x00000000)
Computer ID 7952666
Run time 7 min 2 sec
CPU time 1 min 8 sec
Validate state Invalid
Credit 0.00
Device peak FLOPS 1,862.40 GFLOPS
Application version SETI@home v8
Anonymous platform (NVIDIA GPU)
Peak working set size 500.03 MB
Peak swap size 82,839.78 MB
Peak disk usage 0.14 MB


Stderr output

<core_client_version>7.8.6</core_client_version>
<![CDATA[
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1: GeForce GTX 1050, 2047 MiB, regsPerBlock 65536
computeCap 6.1, multiProcs 5
pciBusID = 3, pciSlotID = 0
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 1050 is okay
SETI@home using CUDA accelerated device GeForce GTX 1050
Using unroll = 5 from command line args

setiathome v8 enhanced x41p_V0.97b2, Cuda 9.10 special
Modifications done by petri33, compiled by TBar

Detected setiathome_enhanced_v8 task. Autocorrelations enabled, size 128k elements.
Work Unit Info:
...............
WU true angle range is : 0.013610
Sigma 100
Sigma > GaussTOffsetStop: 100 > -36
Thread call stack limit is: 1k
Spike: peak=24.25755, time=77.18, d_freq=1418868756.31, chirp=-6.248, fft_len=64k
Spike: peak=24.23805, time=77.18, d_freq=1418868756.32, chirp=-6.2517, fft_len=64k
Triplet: peak=10.39154, time=11.75, period=9.218, d_freq=1418869528.65, chirp=-8.53, fft_len=16
Triplet: peak=10.71862, time=63.86, period=27.07, d_freq=1418869918.13, chirp=-12.795, fft_len=256
Triplet: peak=11.51033, time=60.23, period=20.26, d_freq=1418869844.3, chirp=-17.325, fft_len=256
Triplet: peak=12.07219, time=60.23, period=20.26, d_freq=1418869834.24, chirp=-18.126, fft_len=128
Find triplets Cuda kernel encountered too many triplets, or bins above threshold, reprocessing this PoT on CPU... err = 1

Best spike: peak=24.25755, time=77.18, d_freq=1418868756.31, chirp=-6.248, fft_len=64k
Best autocorr: peak=17.51225, time=100.7, delay=2.9309, d_freq=1418869354.38, chirp=21.526, fft_len=128k
Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.124e+11, d_freq=0,
score=-12, null_hyp=0, chirp=0, fft_len=0
Best pulse: peak=0, time=-2.124e+11, period=0, d_freq=0, score=0, chirp=0, fft_len=0
Best triplet: peak=12.07219, time=60.23, period=20.26, d_freq=1418869834.24, chirp=-18.126, fft_len=128

Spike count: 2
Autocorr count: 0
Pulse count: 0
Triplet count: 4
Gaussian count: 0

04:25:59 (30973): called boinc_finish(0)

</stderr_txt>
]]>

------------------------------- End StdErr Report ----------------------------------

Just rebooted the Computer. Let's see if this makes a difference.


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969096 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969100 - Posted: 7 Dec 2018, 16:30:07 UTC
Last modified: 7 Dec 2018, 16:37:00 UTC

[Event Log AFTER Reboot:]

Fri Dec 7 07:48:45 2018 | | cc_config.xml not found - using defaults
Fri Dec 7 07:48:46 2018 | | Starting BOINC client version 7.8.6 for x86_64-apple-darwin
Fri Dec 7 07:48:46 2018 | | log flags: file_xfer, sched_ops, task
Fri Dec 7 07:48:46 2018 | | Libraries: libcurl/7.50.2 OpenSSL/1.1.0 zlib/1.2.11 c-ares/1.11.0
Fri Dec 7 07:48:46 2018 | | Data directory: /Library/Application Support/BOINC Data
Fri Dec 7 07:48:47 2018 | | CUDA: NVIDIA GPU 0: GeForce GTX 1050 (driver version 10.0.130, CUDA version 10.0, compute capability 6.1, 2048MB, 1408MB available, 1862 GFLOPS peak)
Fri Dec 7 07:48:47 2018 | | OpenCL: NVIDIA GPU 0: GeForce GTX 1050 (driver version 10.33.0 387.10.10.10.40.108, device version OpenCL 1.2, 2048MB, 1408MB available, 1862 GFLOPS peak)
Fri Dec 7 07:48:47 2018 | | OpenCL: AMD/ATI GPU 0: ATI Radeon HD 5770 (driver version 1.2 (Oct 10 2018 23:57:11), device version OpenCL 1.2, 1024MB, 1024MB available, 680 GFLOPS peak)
Fri Dec 7 07:48:47 2018 | | OpenCL CPU: Intel(R) Xeon(R) CPU E5620 @ 2.40GHz (OpenCL driver vendor: Apple, driver version 1.1, device version OpenCL 1.2)
Fri Dec 7 07:48:47 2018 | SETI@home | Found app_info.xml; using anonymous platform
Fri Dec 7 07:48:47 2018 | | Host name: Andromeda.local
Fri Dec 7 07:48:47 2018 | | Processor: 16 GenuineIntel Intel(R) Xeon(R) CPU E5620 @ 2.40GHz [x86 Family 6 Model 44 Stepping 2]
Fri Dec 7 07:48:47 2018 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clfsh ds acpi mmx fxsr sse sse2 ss htt tm pbe pni pclmulqdq dtes64 mon dscpl vmx smx est tm2 ssse3 cx16 tpr pdcm sse4_1 sse4_2 popcnt aes pcid
Fri Dec 7 07:48:47 2018 | | OS: Mac OS X 10.13.6 (Darwin 17.7.0)
Fri Dec 7 07:48:47 2018 | | Memory: 64.00 GB physical, 653.94 GB virtual
Fri Dec 7 07:48:47 2018 | | Disk: 930.71 GB total, 653.69 GB free
Fri Dec 7 07:48:47 2018 | | Local time is UTC -8 hours
Fri Dec 7 07:48:47 2018 | | VirtualBox version: 5.2.18r124319
Fri Dec 7 07:48:47 2018 | | A new version of BOINC is available (7.14.2). <a href=https://boinc.berkeley.edu/download.php>Download</a>
Fri Dec 7 07:48:47 2018 | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 7952666; resource share 365
Fri Dec 7 07:48:47 2018 | SETI@home | General prefs: from SETI@home (last modified 04-May-2018 09:14:36)
Fri Dec 7 07:48:47 2018 | SETI@home | Computer location: home
Fri Dec 7 07:48:47 2018 | SETI@home | General prefs: no separate prefs for home; using your defaults
Fri Dec 7 07:48:47 2018 | | Reading preferences override file
Fri Dec 7 07:48:47 2018 | | Preferences:
Fri Dec 7 07:48:47 2018 | | max memory usage when active: 22937.60 MB
Fri Dec 7 07:48:47 2018 | | max memory usage when idle: 32768.00 MB
Fri Dec 7 07:48:47 2018 | | max disk usage: 30.00 GB
Fri Dec 7 07:48:47 2018 | | max CPUs used: 8
Fri Dec 7 07:48:47 2018 | | (to change preferences, visit a project web site or select Preferences in the Manager)
Fri Dec 7 07:48:47 2018 | | Running CPU benchmarks
Fri Dec 7 07:48:47 2018 | | Suspending computation - CPU benchmarks in progress
Fri Dec 7 07:49:18 2018 | | Benchmark results:
Fri Dec 7 07:49:18 2018 | | Number of CPUs: 8
Fri Dec 7 07:49:18 2018 | | 3254 floating point MIPS (Whetstone) per CPU
Fri Dec 7 07:49:18 2018 | | 11910 integer MIPS (Dhrystone) per CPU
Fri Dec 7 07:55:58 2018 | SETI@home | project resumed by user
Fri Dec 7 07:56:42 2018 | SETI@home | General prefs: from SETI@home (last modified 04-May-2018 09:14:36)
Fri Dec 7 07:56:42 2018 | SETI@home | Computer location: home
Fri Dec 7 07:56:42 2018 | SETI@home | General prefs: no separate prefs for home; using your defaults
Fri Dec 7 07:56:42 2018 | | Reading preferences override file
Fri Dec 7 07:56:42 2018 | | Preferences:
Fri Dec 7 07:56:42 2018 | | max memory usage when active: 22937.60 MB
Fri Dec 7 07:56:42 2018 | | max memory usage when idle: 32768.00 MB
Fri Dec 7 07:56:42 2018 | | max disk usage: 30.00 GB
Fri Dec 7 07:56:42 2018 | | max CPUs used: 8
Fri Dec 7 07:56:42 2018 | | (to change preferences, visit a project web site or select Preferences in the Manager)

---------------------------- End Event Log. ----------------------------

So, WHY has my Available VRAM DROPPED to 1.4GB??? NO Monitor is Attached!!! Is this why I'm getting the Invalids and Validation Inconclusives???


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969100 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969105 - Posted: 7 Dec 2018, 17:42:01 UTC

Just set her to No New Tasks... :-(


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969105 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969133 - Posted: 7 Dec 2018, 19:34:33 UTC
Last modified: 7 Dec 2018, 19:38:20 UTC

The problem seems to be from 3:47 AM on the 5th to 7:05 PM on the 6th for the V.I.s... The Invalids are all in this time window, too. [EDIT:] (All times are Pacific Time Zone.)

There was an OS glitch on the 5th that I caught where my Apple WiFi Card shut itself completely off, thus blocking BOINC's communication... I turned the WiFi Card back on and it automatically connected to my Network and BOINC began reporting Completed Tasks. I thought everything was OK at that point... Apparently NOT! :-(

I still have BOINC set to NNT. I have quite a few Dec 7th Reported Tasks that ARE Validating, now... Maybe the glitch that shut off my WiFi was more systemic and affected BOINC, and maybe other areas in the OS. Hopefully this morning's reboot has resolved the issue.

If things are back to "normal" now, I DON'T have high hopes of the 32 remaining V.I.s Validating... MOST of them may go Invalid. :-(


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969133 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969183 - Posted: 7 Dec 2018, 23:15:09 UTC

Invalids keep fluctuating. (Going from 6 to 10 and back down to 6.)

Validation Inconclusives are down to 30. (Down from the high of 38.)

I'd still like to know WTF the OS did to BORK all these Units!!!


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969183 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4859
Credit: 582,540,799
RAC: 1,342,675
United States
Message 1969196 - Posted: 8 Dec 2018, 0:08:22 UTC - in response to Message 1969183.  

I'd say something triggered the bug that causes the app to not find pulses. Usually you can avoid that bug by running the monitor on an ATI card. I'd say stop worrying about tasks done over a day ago and concentrate on the current tasks. As long as the current tasks show a Best Pulse in the stderr the task will be good. If it's missing a Best Pulse then you need to restart the machine or just swap the monitor cable from the ATI card to the NV card for about 30 seconds and then move it back to the ATI card. As long as you just have BOINC running you can run the monitor on the NV card for that long. Then check the next task and see if it has a Best Pulse. You also might want to update your App as a New one was posted some time ago, Message 1955220 - Posted: 13 Sep 2018, 17:01:37 UTC What's interesting is this machine has never suffered the Bug, that I'm aware of. For my fast machine, I just swap the cable from one GPU to the next and when they all find pulses I turn the monitor off until the next time I need it.
ID: 1969196 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969202 - Posted: 8 Dec 2018, 0:40:23 UTC
Last modified: 8 Dec 2018, 0:57:46 UTC

TBar,

Just Updated the App. Thought I already had the Latest Ver, (b2), guess not.

Let's see what happens, now.


TL

[EDIT:]

[Pending Validation - New App Recently Reported Unit:]

Task 7218648337

Name blc13_2bit_guppi_58405_85640_HIP85417_0027.19990.409.22.45.9.vlar_0
Workunit 3253270930
Created 7 Dec 2018, 16:07:45 UTC
Sent 7 Dec 2018, 21:40:45 UTC
Report deadline 30 Jan 2019, 2:40:27 UTC
Received 8 Dec 2018, 0:51:28 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x00000000)
Computer ID 7952666
Run time 3 min 59 sec
CPU time 42 sec

Validate state Initial
Credit 0.00
Device peak FLOPS 1,862.40 GFLOPS
Application version SETI@home v8
Anonymous platform (NVIDIA GPU)
Peak working set size 465.99 MB
Peak swap size 82,774.29 MB
Peak disk usage 0.12 MB

Stderr output

<core_client_version>7.8.6</core_client_version>
<![CDATA[
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1: GeForce GTX 1050, 2047 MiB, regsPerBlock 65536
computeCap 6.1, multiProcs 5
pciBusID = 3, pciSlotID = 0
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 1050 is okay
SETI@home using CUDA accelerated device GeForce GTX 1050
Using unroll = 5 from command line args

setiathome v8 enhanced x41p_V0.97, Cuda 9.00 special
Modifications done by petri33, compiled by TBar

Detected setiathome_enhanced_v8 task. Autocorrelations enabled, size 128k elements.
Work Unit Info:
...............
WU true angle range is : 0.005864
Sigma 117
Sigma > GaussTOffsetStop: 117 > -53
Thread call stack limit is: 1k
Triplet: peak=11.54149, time=30.67, period=5.693, d_freq=6571392781.83, chirp=-8.4915, fft_len=128
Autocorr: peak=17.83442, time=28.63, delay=3.5429, d_freq=6571391558.63, chirp=-17.469, fft_len=128k
Pulse: peak=3.65578, time=45.9, period=7.896, d_freq=6571393707.17, score=1.002, chirp=47.231, fft_len=2k
Pulse: peak=5.748226, time=45.84, period=14.94, d_freq=6571394589.78, score=1.003, chirp=-82.787, fft_len=512

Best spike: peak=23.22259, time=82.32, d_freq=6571394666.58, chirp=38.992, fft_len=16k
Best autocorr: peak=17.83442, time=28.63, delay=3.5429, d_freq=6571391558.63, chirp=-17.469, fft_len=128k
Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.124e+11, d_freq=0,
score=-12, null_hyp=0, chirp=0, fft_len=0
Best pulse: peak=5.748226, time=45.84, period=14.94, d_freq=6571394589.78, score=1.003, chirp=-82.787, fft_len=512
Best triplet: peak=11.54149, time=30.67, period=5.693, d_freq=6571392781.83, chirp=-8.4915, fft_len=128

Spike count: 0
Autocorr count: 1
Pulse count: 2
Triplet count: 1
Gaussian count: 0

16:47:01 (5562): called boinc_finish(0)

</stderr_txt>
]]>
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969202 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969271 - Posted: 8 Dec 2018, 7:06:43 UTC

Just Updated to HS Sec Update 2018-003 and Web Driver 387.10.10.10.40.113.

New Event Log:

Fri Dec 7 21:38:00 2018 | | cc_config.xml not found - using defaults
Fri Dec 7 21:38:00 2018 | | Starting BOINC client version 7.8.6 for x86_64-apple-darwin
Fri Dec 7 21:38:00 2018 | | log flags: file_xfer, sched_ops, task
Fri Dec 7 21:38:00 2018 | | Libraries: libcurl/7.50.2 OpenSSL/1.1.0 zlib/1.2.11 c-ares/1.11.0
Fri Dec 7 21:38:00 2018 | | Data directory: /Library/Application Support/BOINC Data
Fri Dec 7 21:38:01 2018 | | CUDA: NVIDIA GPU 0: GeForce GTX 1050 (driver version 10.0.130, CUDA version 10.0, compute capability 6.1, 2048MB, 1343MB available, 1862 GFLOPS peak)
Fri Dec 7 21:38:01 2018 | | OpenCL: NVIDIA GPU 0: GeForce GTX 1050 (driver version 10.33.0 387.10.10.10.40.113, device version OpenCL 1.2, 2048MB, 1343MB available, 1862 GFLOPS peak)

Fri Dec 7 21:38:01 2018 | | OpenCL: AMD/ATI GPU 0: ATI Radeon HD 5770 (driver version 1.2 (Nov 4 2018 20:14:52), device version OpenCL 1.2, 1024MB, 1024MB available, 680 GFLOPS peak)
Fri Dec 7 21:38:01 2018 | | OpenCL CPU: Intel(R) Xeon(R) CPU E5620 @ 2.40GHz (OpenCL driver vendor: Apple, driver version 1.1, device version OpenCL 1.2)
Fri Dec 7 21:38:01 2018 | SETI@home | Found app_info.xml; using anonymous platform
Fri Dec 7 21:38:01 2018 | | Host name: Andromeda.local
Fri Dec 7 21:38:01 2018 | | Processor: 16 GenuineIntel Intel(R) Xeon(R) CPU E5620 @ 2.40GHz [x86 Family 6 Model 44 Stepping 2]
Fri Dec 7 21:38:01 2018 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clfsh ds acpi mmx fxsr sse sse2 ss htt tm pbe pni pclmulqdq dtes64 mon dscpl vmx smx est tm2 ssse3 cx16 tpr pdcm sse4_1 sse4_2 popcnt aes pcid
Fri Dec 7 21:38:01 2018 | | OS: Mac OS X 10.13.6 (Darwin 17.7.0)
Fri Dec 7 21:38:01 2018 | | Memory: 64.00 GB physical, 656.06 GB virtual
Fri Dec 7 21:38:01 2018 | | Disk: 930.71 GB total, 655.81 GB free
Fri Dec 7 21:38:01 2018 | | Local time is UTC -8 hours
Fri Dec 7 21:38:01 2018 | | VirtualBox version: 5.2.18r124319
Fri Dec 7 21:38:01 2018 | | A new version of BOINC is available (7.14.2). <a href=https://boinc.berkeley.edu/download.php>Download</a>
Fri Dec 7 21:38:01 2018 | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 7952666; resource share 365
Fri Dec 7 21:38:01 2018 | SETI@home | General prefs: from SETI@home (last modified 04-May-2018 09:14:36)
Fri Dec 7 21:38:01 2018 | SETI@home | Computer location: home
Fri Dec 7 21:38:01 2018 | SETI@home | General prefs: no separate prefs for home; using your defaults
Fri Dec 7 21:38:01 2018 | | Reading preferences override file
Fri Dec 7 21:38:01 2018 | | Preferences:
Fri Dec 7 21:38:01 2018 | | max memory usage when active: 22937.60 MB
Fri Dec 7 21:38:01 2018 | | max memory usage when idle: 32768.00 MB
Fri Dec 7 21:38:01 2018 | | max disk usage: 30.00 GB
Fri Dec 7 21:38:01 2018 | | max CPUs used: 8
Fri Dec 7 21:38:01 2018 | | (to change preferences, visit a project web site or select Preferences in the Manager)

-------------------------- End Event Log. ----------------------------

NOTICE: NOW, Available VRAM on the 1050 is 1343MB DOWN AGAIN!!! On 387.10.10.10.40.108 and Sec Updt 002, Available VRAM was 1408MB. AND, I believe on the Prior Driver to that, and Pre-Sec Updt 002, the VRAM Available was MUCH higher...

STILL NO Monitor attached to the 1050. ONLY the ATI HD5770 has a Monitor attached - the Apple Cinema Display - 23". So, WHY is SOOOO MUCH VRAM being taken from the 1050?????


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969271 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969334 - Posted: 8 Dec 2018, 15:36:41 UTC
Last modified: 8 Dec 2018, 15:41:02 UTC

TBar,

Unfortunately, the Bug that I've hit in your App is too prevalent for me... It seems that, now, Crunching at night, (like I usually have my computer do), with the monitor off - since the 5th - is NO LONGER a "doable" thing. It seems that with the monitor off at night that this "Best Pulse = 0" Bug hit MANY Units during last night's Crunching, too.

I've set the Computer to NNT, and I'm Crunching through my Queue, and then WON'T Crunch anymore on the Mac! I CANNOT leave the Apple Cinema Display running 24/7 as it seems to heat up in the upper left corner quite a bit while on for more than an hour. (I notice this effect on iMacs, too. My sister's old iMac does this all the time.) I have no money to replace the monitor, so I guess - no more Crunching is in my future on Mac. :-(

[EDIT:]

BUT, why turning the monitor off on the ATI Card is making the NVIDIA Card hit "Best Pulse = 0" is a COMPLETE mystery to me... :-O


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969334 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2766
Credit: 568,338,037
RAC: 882,980
Canada
Message 1969335 - Posted: 8 Dec 2018, 15:46:22 UTC - in response to Message 1969334.  

You could get a dummy HDMI plug and plug that in when you are away.
If you don't mind swapping cables ...
ID: 1969335 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1969347 - Posted: 8 Dec 2018, 16:51:28 UTC - in response to Message 1969335.  

You could get a dummy HDMI plug and plug that in when you are away.
If you don't mind swapping cables ...

How would this work? The ATI HD5770 has NO HDMI Port, it's too old of a card. It has 2 Mini-Display Ports and one DVI Port. The Cinema Display - 23" is on DVI.

The NVIDIA Card has one HDMI. Are you saying that I need to plug a dummy HDMI Plug into the NVIDIA Card??? If this is the case, what's the difference, (load wise), between having a Dummy Plug vs a Monitor on the NVIDIA Card? I ask because TBar says that NO Monitor should be on my 1050 Card as it ONLY has 2GB VRAM. So, if the Dummy Plug puts a "fake" Monitor Load, that would cause Units to crash and fail, too...


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1969347 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4859
Credit: 582,540,799
RAC: 1,342,675
United States
Message 1969357 - Posted: 8 Dec 2018, 17:41:01 UTC

Well, it does seem to be the monitor changing state that triggers the missing pulses. If you can stop that it would probably work OK. I have 4 Macs, well 1 Mac and 3 Hacks, on two of them making sure the App is not missing pulses and then turning off the monitor works, until you turn the monitor back on and start using the machine. On the real Mac, it works fine as long as the monitor connected to the ATI card is running, but seems to start missing pulses when the monitor goes to sleep. The Fourth machine is in the TV room connected to a TV that stays on all the time and is connected by another ATI card. The one connected to the TV seems to work fine too, but, the TV stays on 24/7. The real Mac was working fine last night when I went to sleep, but after the monitor went to sleep it started missing pulses, this morning I simply changed the monitor cable over to the NV card and then back to the ATI card and now it is working normally again. So, only crunch when you can make sure the monitor state doesn't change. For me, I'm just going to use the 5 card Hack sitting over in the corner with the monitor off. So, Petri blames the NV drivers, and NV will undoubtedly blame Petri's App as they did the last time a bug report was filed, looks like a no win situation. Apple does somewhat side with Petri though, Apple doesn't care much for the recent NV drivers either.
ID: 1969357 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1971005 - Posted: 19 Dec 2018, 3:01:09 UTC

Thanks to Carlos, I now have a Monitor I can leave on, (with Apple Screen Saver of course), and NO LONGER encounter the CUDA90 App Bug. :-)

The Computer is now on a Viewsonic VX2250, 21.5" LED/LCD Monitor. Supports 1920 x 1080. It's been on ALL day today, (12-18-2018), and DOES NOT get hot like the 23" Apple Cinema Display. :-)

Monitor, (again), connected to the ATI HD-5770 GPU. The ATI Card DOES NOT Crunch. ONLY the GTX-1050 Crunches.

Let's see if I can get my RAC back up above 13.5K. :-)


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1971005 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9839
Credit: 922,922,234
RAC: 1,520,212
United States
Message 1971011 - Posted: 19 Dec 2018, 3:54:58 UTC - in response to Message 1969357.  

All it takes is the DVI to VGA adapter that is included with all Nvidia Pascal cards and three 1/8W 75 ohm resistors to make a "dummy" VGA plug that fools the card into thinking a monitor is connected.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1971011 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 62543
Credit: 51,149,891
RAC: 20,155
United States
Message 1971024 - Posted: 19 Dec 2018, 7:00:18 UTC - in response to Message 1971005.  

Thanks to Carlos, I now have a Monitor I can leave on, (with Apple Screen Saver of course), and NO LONGER encounter the CUDA90 App Bug. :-)

The Computer is now on a Viewsonic VX2250, 21.5" LED/LCD Monitor. Supports 1920 x 1080. It's been on ALL day today, (12-18-2018), and DOES NOT get hot like the 23" Apple Cinema Display. :-)

Monitor, (again), connected to the ATI HD-5770 GPU. The ATI Card DOES NOT Crunch. ONLY the GTX-1050 Crunches.

Let's see if I can get my RAC back up above 13.5K. :-)


TL

TL, there is such a thing as a Dummy Plug for DP and Mini DP, I found a bunch on Amazon. Though some come in 3's and some don't.
My Amazon Wishlist
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, One of America's First HST's
ID: 1971024 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1971044 - Posted: 19 Dec 2018, 14:02:54 UTC

Guys,

I use this computer daily. Why would I want a dummy plug that I'd have to keep changing back out for the monitor????

The System is now where I need it to be... It has a Monitor that I can leave on and does NOT get hot like the Apple Cinema Display does.


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1971044 · Report as offensive
Profile TimeLord04 Project Donor
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 19953
Credit: 25,326,140
RAC: 63,469
United States
Message 1971141 - Posted: 20 Dec 2018, 3:19:38 UTC

Deeply disappointed... :-(

First Unit done for the day - turns Invalid!

-------------------------------------------------------

Task 7252993560

Name 16dc18ad.10819.169987.9.36.0_1
Workunit 3269886760

Created 18 Dec 2018, 20:53:08 UTC
Sent 18 Dec 2018, 23:41:29 UTC
Report deadline 10 Feb 2019, 0:12:52 UTC
Received 19 Dec 2018, 5:15:42 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x00000000)
Computer ID 7952666
Run time 5 sec
CPU time 3 sec
Validate state Invalid

Credit 0.00
Device peak FLOPS 1,862.40 GFLOPS
Application version SETI@home v8
Anonymous platform (NVIDIA GPU)
Peak disk usage 0.20 MB

Stderr output

<core_client_version>7.8.6</core_client_version>
<![CDATA[
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1: GeForce GTX 1050, 2047 MiB, regsPerBlock 65536
computeCap 6.1, multiProcs 5
pciBusID = 3, pciSlotID = 0
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 1050 is okay
SETI@home using CUDA accelerated device GeForce GTX 1050
Using unroll = 5 from command line args

setiathome v8 enhanced x41p_V0.97, Cuda 9.00 special
Modifications done by petri33, compiled by TBar

Detected setiathome_enhanced_v8 task. Autocorrelations enabled, size 128k elements.
Work Unit Info:
...............
WU true angle range is : 0.432616
Sigma 3
Thread call stack limit is: 1k
Find triplets Cuda kernel encountered too many triplets, or bins above threshold, reprocessing this PoT on CPU... err = 1
Triplet: peak=10.37088, time=0.8532, period=0.08847, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.5631, time=4.761, period=2.277, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=9.523264, time=7.094, period=4.115, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=11.50822, time=4.962, period=0.4145, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=11.77172, time=7.038, period=2.277, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.66159, time=7.757, period=2.795, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.68646, time=7.757, period=1.262, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.40651, time=10.55, period=0.6218, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.64058, time=7.757, period=1.262, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.36183, time=10.55, period=0.6218, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.53419, time=16.67, period=1.036, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.58566, time=16.67, period=1.036, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.42603, time=22.51, period=0.6218, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.43884, time=22.98, period=0.4723, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.43884, time=23.39, period=0.8868, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.57415, time=22.97, period=0.4145, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.58154, time=23.18, period=0.6218, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.23921, time=23.5, period=0.777, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.58154, time=23.18, period=0.2073, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.57415, time=23.39, period=0.4145, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.32101, time=23.39, period=0.4145, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.226, time=23.81, period=0.829, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.58154, time=23.39, period=0.2073, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.57415, time=23.59, period=0.2073, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=11.53342, time=24.02, period=0.6218, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.97647, time=23.71, period=0.2073, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.55425, time=22.51, period=0.6218, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.54656, time=22.98, period=0.4723, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=10.54656, time=23.39, period=0.8868, d_freq=1419995117.19, chirp=0, fft_len=8
Triplet: peak=12.7039, time=22.97, period=0.4145, d_freq=1419995117.19, chirp=0, fft_len=8
SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected equals the storage space allocated. <<----- What's THIS???!!!???

Best spike: peak=7.999279, time=102.3, d_freq=1419995117.19, chirp=0, fft_len=8
Best autocorr: peak=0, time=-2.124e+11, delay=0, d_freq=0, chirp=0, fft_len=0
Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.124e+11, d_freq=0,
score=-12, null_hyp=0, chirp=0, fft_len=0
Best pulse: peak=0, time=-2.124e+11, period=0, d_freq=0, score=0, chirp=0, fft_len=0
Best triplet: peak=12.7039, time=22.97, period=0.4145, d_freq=1419995117.19, chirp=0, fft_len=8

Spike count: 0
Autocorr count: 0
Pulse count: 0
Triplet count: 30
Gaussian count: 0

21:10:08 (18185): called boinc_finish(0)

</stderr_txt>
]]>


-----------------------------------------------

Viewsonic Monitor ON all day, connected SOLELY to the ATI Card. So WTF?????


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1971141 · Report as offensive
Kissagogo27 Special Project $75 donor

Send message
Joined: 6 Nov 99
Posts: 513
Credit: 7,008,740
RAC: 3,968
France
Message 1971200 - Posted: 20 Dec 2018, 14:50:11 UTC

like the others results, it's a noisy wu

3269886760
ID: 1971200 · Report as offensive
1 · 2 · Next

Message boards : Number crunching : Validation Inconclusive - Mac CUDA90 App - 38 V.I.!!!


 
©2019 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.