Open Beta test: SoG for NVidia, Lunatics v0.45 - Beta6 (RC again)

Message boards : Number crunching : Open Beta test: SoG for NVidia, Lunatics v0.45 - Beta6 (RC again)
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 24 · 25 · 26 · 27 · 28 · 29 · 30 . . . 32 · Next

AuthorMessage
Profile Cliff Harding
Volunteer tester
Avatar

Send message
Joined: 18 Aug 99
Posts: 1432
Credit: 110,967,840
RAC: 67
United States
Message 1829204 - Posted: 8 Nov 2016, 14:52:14 UTC - in response to Message 1829188.  

Downloaded and installed -6 successfully. Will report back after I'VE VOTED!!


I don't buy computers, I build them!!
ID: 1829204 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13159
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1829226 - Posted: 9 Nov 2016, 3:02:00 UTC

Had one computer complain about missing the AP 2887 application being missing in app_config even though the app was installed in the directory. Ran the Lunatics installer again and all cleared up. Might have been too fast mousing or something. No harm, no foul. All good now.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1829226 · Report as offensive
robertmiles
Volunteer tester

Send message
Joined: 16 Jan 12
Posts: 213
Credit: 4,117,756
RAC: 6
United States
Message 1829231 - Posted: 9 Nov 2016, 3:28:05 UTC

A minor problem with the beta 6 installation under 64-bit Windows Vista: At the end, the front window of the two did not close until I clicked Finish in the rear window.

Not enough tasks finished yet to tell if there is any other problem, but the first three looked successful.
ID: 1829231 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13415
Credit: 208,696,464
RAC: 304
Australia
Message 1829234 - Posted: 9 Nov 2016, 3:42:16 UTC - in response to Message 1829231.  

Ran & installed SoG with no problems, cache still intact.
Grant
Darwin NT
ID: 1829234 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1829261 - Posted: 9 Nov 2016, 6:52:06 UTC - in response to Message 1829188.  

OK, after yet another new build on Sunday night, we seem to have a brief pause between server crashes, so I've uploaded...

Beta 6

The key features are:
NV SoG r3557, aimed at reducing the number of inconclusive validations
Intel GPU r3541 (that's the same as v8.20 at Beta), which solves the long-standing problem of validation on Skylake 6th. generation iGPUs

[BTW, has anyone ever tried running Astropulse on a Skylake?]

Every time I look at OneDrive, it shows me a different url, but the one we used last time still seems to work:

https://onedrive.live.com/redir?resid=8D83BF774A4A86F5!979&authkey=!AK7t-36P9nAec2Y&ithint=folder%2cexe

As usual, expect your antivirus program to be suspicious of the new files until a number of you have downloaded them.


. . Hi Richard,

. . Is there much point in installing Beta6 if I am not having problems with Beta5??

Stephen

.
ID: 1829261 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6324
Credit: 106,370,077
RAC: 121
Russia
Message 1829289 - Posted: 9 Nov 2016, 10:31:34 UTC - in response to Message 1829261.  

. . Is there much point in installing Beta6 if I am not having problems with Beta5??

.

If you use any of OpenCL SoG apps - yes, you need update.
SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1829289 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1829291 - Posted: 9 Nov 2016, 10:47:40 UTC - in response to Message 1829289.  

. . Is there much point in installing Beta6 if I am not having problems with Beta5??

.

If you use any of OpenCL SoG apps - yes, you need update.


. . Hi Raistmer,

. . Thanks for that, I will download it shortly.

Stephen

.
ID: 1829291 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1829313 - Posted: 9 Nov 2016, 12:46:00 UTC - in response to Message 1829188.  

OK, after yet another new build on Sunday night, we seem to have a brief pause between server crashes, so I've uploaded...

Beta 6

The key features are:
NV SoG r3557, aimed at reducing the number of inconclusive validations
Intel GPU r3541 (that's the same as v8.20 at Beta), which solves the long-standing problem of validation on Skylake 6th. generation iGPUs

[BTW, has anyone ever tried running Astropulse on a Skylake?]

Every time I look at OneDrive, it shows me a different url, but the one we used last time still seems to work:

https://onedrive.live.com/redir?resid=8D83BF774A4A86F5!979&authkey=!AK7t-36P9nAec2Y&ithint=folder%2cexe

As usual, expect your antivirus program to be suspicious of the new files until a number of you have downloaded them.



. . Hi Richard,

. . If I remember back in Feb/Mar when I bought the i5 it was not very successful running Astropulses but I think that came good when I updated to later (beta) video drivers (I have since added a GTX950 and cannot remember what drivers they were, hang on, I may still have the download - Beta_Intel Video 15.40.18.64.4380). But I ceased using the Intel GPU not long after that because I fould using it choked up the CPU potential. What little gain came from the iGPU was more greatly offset by the drop in output from crunching on the CPU and I more than doubled my output using just the CPU cores and Lunatics/AVX.

. . I have installed Beta6 and had a little scare; I ran it with Boinc running and on the restart it came up with no tasks (or project) so I had to close it and restart it manually and then it came up AOK. Now we will see if it makes any difference but that is for the support thread.

Stephen

.
ID: 1829313 · 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: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1829317 - Posted: 9 Nov 2016, 12:52:54 UTC - in response to Message 1829313.  

Astropulse should be fine on your 750 and 950 with your driver 372.
ID: 1829317 · Report as offensive
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 1282
Credit: 187,688,550
RAC: 182
United States
Message 1829326 - Posted: 9 Nov 2016, 14:50:33 UTC - in response to Message 1829317.  
Last modified: 9 Nov 2016, 14:50:47 UTC

Downloaded and now returning 3557 processed WUs.
Only thing I noticed: on the first computer I exited BOINC before I started the install and all went well. On the second computer I decided to let the installer shut down BOINC. I am not sure what happened but it did not close BLOINC and I had to end up doing the install again and this time I exited BOINC before I ran the installer.
I am still getting inconclusives but I will have to watch it a bit longer to see if the number goes down.

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1829326 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1829332 - Posted: 9 Nov 2016, 15:25:07 UTC - in response to Message 1829317.  

Astropulse should be fine on your 750 and 950 with your driver 372.



. . Hi Brent,

. . No problems with APs on any of my nvidia cards with any of the current drivers, but Richard was asking about how they ran on the Intel GPUs (such as the HD530 built into my i5-6400). I think they will run OK with drivers later than Feb/Mar this year.

Stephen

.
ID: 1829332 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1829334 - Posted: 9 Nov 2016, 15:32:51 UTC - in response to Message 1829326.  
Last modified: 9 Nov 2016, 15:34:39 UTC

Downloaded and now returning 3557 processed WUs.
Only thing I noticed: on the first computer I exited BOINC before I started the install and all went well. On the second computer I decided to let the installer shut down BOINC. I am not sure what happened but it did not close BLOINC and I had to end up doing the install again and this time I exited BOINC before I ran the installer.
I am still getting inconclusives but I will have to watch it a bit longer to see if the number goes down.


. . Hi Bill

. . I had a similar problem, Boinc did not close properly but the installation ran anyway. When it should have restarted boinc_tray (and presumably Boinc client) seemed to be in limbo and Boinc Manager would not talk to them. Rather than rush into a second instal attempt I manually closed and restarted Boinc and this time it initialised AOK and everything seemed to be as it should. I do think it would be wise to manually exit Boinc before running the setup app. I cannot tell at this point whether any tasks were ghosted but it seems not.

Stephen

PS: I have returned only two dozen results so far so it is too early to notice any pattern in the validation outcomes.

Stephen

.
ID: 1829334 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14542
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1829349 - Posted: 9 Nov 2016, 16:39:25 UTC - in response to Message 1829334.  

. . I had a similar problem, Boinc did not close properly but the installation ran anyway. When it should have restarted boinc_tray (and presumably Boinc client) seemed to be in limbo and Boinc Manager would not talk to them. Rather than rush into a second instal attempt I manually closed and restarted Boinc and this time it initialised AOK and everything seemed to be as it should. I do think it would be wise to manually exit Boinc before running the setup app. I cannot tell at this point whether any tasks were ghosted but it seems not.

For clarity and accuracy:

The installer is designed to stop (if necessary) and restart (if stopped) the boinc client only - that's boinc.exe

It doesn't touch boinctray.exe (the invisible user activity detection tool) or boincmgr.exe (BOINC Manager, the source of the visible icon in the notification area).

The best guess I have is that on modern, fast machines the interval between stopping boinc.exe and restarting boinc.exe - the period when all the installer's work is done - is simply too short: something prevents boinc.exe starting again that soon. By the time you notice the problem, exit the Manager, and restart the Manager again, enough time has elapsed for BOINC Manager to successfully restart boinc.exe - as it's designed to do.

I'll try and slow the installer down a bit before final release.
ID: 1829349 · Report as offensive
Profile Jimbocous Project Donor
Volunteer tester
Avatar

Send message
Joined: 1 Apr 13
Posts: 1849
Credit: 268,616,081
RAC: 1,349
United States
Message 1829412 - Posted: 9 Nov 2016, 20:09:53 UTC - in response to Message 1829349.  
Last modified: 9 Nov 2016, 20:13:41 UTC

The best guess I have is that on modern, fast machines the interval between stopping boinc.exe and restarting boinc.exe - the period when all the installer's work is done - is simply too short: something prevents boinc.exe starting again that soon. By the time you notice the problem, exit the Manager, and restart the Manager again, enough time has elapsed for BOINC Manager to successfully restart boinc.exe - as it's designed to do.

I'll try and slow the installer down a bit before final release.

Hi, Richard, somewhat related (perhaps) update.
I extensively tested adding delay to the restart of boinc.exe in the QOpt script. Tested 5 hosts 6x per day each for 12 days, with restart delays of 0,5,10 and 15 seconds. I was never able to establish a correlation between restart delay and frequency of restart failure.
I am now testing forcing shutdown of BOINCManager as well, and over 8 days so far have not yet seen a failure to reattach.
But I do agree that relative speed of the machine could be a factor, if only that the faster boxes here were more likely to see the reattach failure.
If you do introduce a restart delay into the installer, I'd be very interested in the results of any testing on this.
Regards, Jim ...
ID: 1829412 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6533
Credit: 196,805,888
RAC: 57
United States
Message 1829488 - Posted: 10 Nov 2016, 0:55:50 UTC - in response to Message 1829349.  

. . I had a similar problem, Boinc did not close properly but the installation ran anyway. When it should have restarted boinc_tray (and presumably Boinc client) seemed to be in limbo and Boinc Manager would not talk to them. Rather than rush into a second instal attempt I manually closed and restarted Boinc and this time it initialised AOK and everything seemed to be as it should. I do think it would be wise to manually exit Boinc before running the setup app. I cannot tell at this point whether any tasks were ghosted but it seems not.

For clarity and accuracy:

The installer is designed to stop (if necessary) and restart (if stopped) the boinc client only - that's boinc.exe

It doesn't touch boinctray.exe (the invisible user activity detection tool) or boincmgr.exe (BOINC Manager, the source of the visible icon in the notification area).

The best guess I have is that on modern, fast machines the interval between stopping boinc.exe and restarting boinc.exe - the period when all the installer's work is done - is simply too short: something prevents boinc.exe starting again that soon. By the time you notice the problem, exit the Manager, and restart the Manager again, enough time has elapsed for BOINC Manager to successfully restart boinc.exe - as it's designed to do.

I'll try and slow the installer down a bit before final release.

I have found at least with BOINC 7.4+ it tend to take a bit longer when you need to cycle it. I imagine they may have gone with a more orderly shutdown app process to prevent files errors or such. Once the boinc.exe process terminates I allow another 5-10 seconds before I'll restart it.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the BP6/VP6 User Group today!
ID: 1829488 · Report as offensive
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 1282
Credit: 187,688,550
RAC: 182
United States
Message 1829597 - Posted: 10 Nov 2016, 14:58:32 UTC - in response to Message 1829488.  

Well at least I am not the only one to have seen this hiccup. I normally do shut down BOINC before running the installer, but since I know the option to have the installer do it I thought I would try it.

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1829597 · Report as offensive
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 1282
Credit: 187,688,550
RAC: 182
United States
Message 1829598 - Posted: 10 Nov 2016, 15:01:23 UTC - in response to Message 1829597.  

I have a question about a WU: http://setiathome.berkeley.edu/workunit.php?wuid=2320732915

I know I do not know where to look exactly, but this looks like it should have validated. We both got the same result, at least to my untrained eyes.

Running 3557 against a Darwin computer.

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1829598 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14542
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1829604 - Posted: 10 Nov 2016, 15:28:04 UTC - in response to Message 1829598.  

Looks like it's the pulses again.

(opencl_intel_gpu_sah) x86_64-apple-darwin
Best pulse: peak=6.931435 , time=96.44, period=2.586 , d_freq=1419474418.49, score=0.9659, chirp=-25.435, fft_len=512

r3557 SoG
Best pulse: peak=0.9932134, time=11.99, period=0.1532, d_freq=1419474804.37, score=0.9999, chirp= 71.006, fft_len=128

Although both are below the reporting threshhold, they're wildly different.
ID: 1829604 · Report as offensive
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 1282
Credit: 187,688,550
RAC: 182
United States
Message 1829610 - Posted: 10 Nov 2016, 15:53:58 UTC - in response to Message 1829604.  

Looks like it's the pulses again.

(opencl_intel_gpu_sah) x86_64-apple-darwin
Best pulse: peak=6.931435 , time=96.44, period=2.586 , d_freq=1419474418.49, score=0.9659, chirp=-25.435, fft_len=512

r3557 SoG
Best pulse: peak=0.9932134, time=11.99, period=0.1532, d_freq=1419474804.37, score=0.9999, chirp= 71.006, fft_len=128

Although both are below the reporting threshhold, they're wildly different.


Thanks Richard. I will have to look at the pulses from now on. I just did not know where to look.

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1829610 · Report as offensive
Profile RueiKe Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 14 Feb 16
Posts: 492
Credit: 378,512,430
RAC: 785
Taiwan
Message 1831724 - Posted: 21 Nov 2016, 11:40:27 UTC

I just installed Beta 6 and checked for the affinity issue noticed when troubleshooting my Penta-Nano system. Not sure if an attempt was made to fix it, but it looks like it is still an issue. One of the 5 running MB tasks has affinity set to all cores, while the other 4 are set to unique cores.
GitHub: Ricks-Lab
Instagram: ricks_labs
ID: 1831724 · Report as offensive
Previous · 1 . . . 24 · 25 · 26 · 27 · 28 · 29 · 30 . . . 32 · Next

Message boards : Number crunching : Open Beta test: SoG for NVidia, Lunatics v0.45 - Beta6 (RC again)


 
©2023 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.