PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise

Advanced search

Message boards : Number crunching : Extremely frustrated-WU's keep restarting

Author Message
Profile Blurf
Send message
Joined: 11 Mar 07
Posts: 155
ID: 6349
Credit: 180,396,240
RAC: 0
321 LLR Amethyst: Earned 1,000,000 credits (1,294,823)Cullen LLR Amethyst: Earned 1,000,000 credits (1,112,563)ESP LLR Amethyst: Earned 1,000,000 credits (1,182,696)Generalized Cullen/Woodall LLR Gold: Earned 500,000 credits (607,415)PPS LLR Amethyst: Earned 1,000,000 credits (1,705,178)SoB LLR Turquoise: Earned 5,000,000 credits (5,000,863)SR5 LLR Amethyst: Earned 1,000,000 credits (1,281,452)SGS LLR Ruby: Earned 2,000,000 credits (2,027,182)TPS LLR (retired) Bronze: Earned 10,000 credits (91,882)TRP LLR Gold: Earned 500,000 credits (500,474)Woodall LLR Gold: Earned 500,000 credits (604,446)321 Sieve (suspended) Ruby: Earned 2,000,000 credits (2,525,881)Cullen/Woodall Sieve (suspended) Silver: Earned 100,000 credits (131,607)Generalized Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (3,782,364)PPS Sieve Double Bronze: Earned 100,000,000 credits (139,942,611)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,007,573)TRP Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,480,387)AP 26/27 Ruby: Earned 2,000,000 credits (2,708,810)GFN Jade: Earned 10,000,000 credits (13,402,257)
Message 99071 - Posted: 23 Sep 2016 | 5:57:04 UTC
Last modified: 23 Sep 2016 | 6:08:46 UTC

Hi all...

Not sure what is wrong but a couple of times my new 8-core monster machine has had to restart (updates, whatever)....my SOB tasks restart from the beginning!

Too late for me to look into it now but any thoughts?

Yes I did abandon some tasks the other day-figured they wouldn't finish in time anyways.

Thx
____________

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 906
ID: 370496
Credit: 482,707,308
RAC: 400,633
321 LLR Jade: Earned 10,000,000 credits (10,008,611)Cullen LLR Jade: Earned 10,000,000 credits (10,005,009)ESP LLR Jade: Earned 10,000,000 credits (10,041,747)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (10,000,820)PPS LLR Jade: Earned 10,000,000 credits (10,020,730)PSP LLR Jade: Earned 10,000,000 credits (10,049,767)SoB LLR Sapphire: Earned 20,000,000 credits (23,095,473)SR5 LLR Jade: Earned 10,000,000 credits (10,003,746)SGS LLR Jade: Earned 10,000,000 credits (10,002,215)TRP LLR Jade: Earned 10,000,000 credits (10,011,903)Woodall LLR Jade: Earned 10,000,000 credits (10,076,850)321 Sieve (suspended) Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Emerald: Earned 50,000,000 credits (50,015,953)GFN Emerald: Earned 50,000,000 credits (53,494,093)WW Emerald: Earned 50,000,000 credits (50,712,000)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 99074 - Posted: 23 Sep 2016 | 6:31:27 UTC - in response to Message 99071.

Hi all...

Not sure what is wrong but a couple of times my new 8-core monster machine has had to restart (updates, whatever)....my SOB tasks restart from the beginning!

Too late for me to look into it now but any thoughts?

Yes I did abandon some tasks the other day-figured they wouldn't finish in time anyways.

Thx

For starters, that "8core" machine is not actually an octa core. For the purposes of Primegrid (specifically LLR tests), which rellies heavilly on floating point calculations, your CPU is actually a quad core, not an octa. The FX lineup had a shared FPU unit across 2 cores, so it only really counts as 4 cores for us, not 8. I don't know if you were running 8 tasks at once, but assuming you were, you are just making them twice as long, if you compare it to running only 4 at a time.

To make things worse, AMD's AVX implementation (which would give tasks a great speedup) is not as good as Intel's, to say the least. It gets to the point that where intel has a big performance boost in the neiborhood of tenths of percents, AMD gets next to no boost at all. In other words, those already long tasks become even longer.

And to top the cherry, the IPC isn't that great either. Aka even more longer to run.


Add all of that up with the fact that SoB tasks aren't just big: they are HUGE. They are REALLY big. The program only checkpoints every so often, so if your PC can't get work done quickly, it'll just go to waste. That's probably why it feels like the tasks are restarting.

If I was you, I'd dedicate that machine to either do TRP-SV, in which it works like a true 8 core, or cunch the smaller LLR tests such as PPS, SR5 or TRP.

DoctorNowProject donor
Avatar
Send message
Joined: 9 Jan 06
Posts: 69
ID: 2121
Credit: 132,329,233
RAC: 41,555
321 LLR Amethyst: Earned 1,000,000 credits (1,026,737)Cullen LLR Amethyst: Earned 1,000,000 credits (1,059,465)ESP LLR Amethyst: Earned 1,000,000 credits (1,049,099)Generalized Cullen/Woodall LLR Gold: Earned 500,000 credits (513,115)PPS LLR Gold: Earned 500,000 credits (891,735)PSP LLR Gold: Earned 500,000 credits (717,680)SoB LLR Gold: Earned 500,000 credits (746,589)SR5 LLR Gold: Earned 500,000 credits (509,394)SGS LLR Gold: Earned 500,000 credits (568,607)TRP LLR Gold: Earned 500,000 credits (560,267)Woodall LLR Gold: Earned 500,000 credits (513,367)321 Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,044,369)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,280,178)Generalized Cullen/Woodall Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,027,256)PPS Sieve Sapphire: Earned 20,000,000 credits (20,156,175)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,153,418)TRP Sieve (suspended) Gold: Earned 500,000 credits (502,748)AP 26/27 Sapphire: Earned 20,000,000 credits (20,161,205)GFN Sapphire: Earned 20,000,000 credits (25,683,733)WW Sapphire: Earned 20,000,000 credits (20,724,000)PSA Sapphire: Earned 20,000,000 credits (29,444,813)
Message 99225 - Posted: 28 Sep 2016 | 8:16:03 UTC - in response to Message 99071.
Last modified: 28 Sep 2016 | 8:24:42 UTC

Not sure what is wrong but a couple of times my new 8-core monster machine has had to restart (updates, whatever)....my SOB tasks restart from the beginning!

Well, I recently bought a similar processor (AMD FX 8320) and didn't know about the bad AVX structure myself, somewhere here in the forum there's already a thread of me about it. ;-)
As per advice I now limit BOINc to use half the cores when crunching LLR tasks, and it works pretty well so far, here are some of my results (hopefully visible to you), these were the recent SoBs I did and I finally reached the gold badge with it.
They are of course not that fast done as when an Intel would do it, but I did find the results satisfactory to me.
Not sure why your tasks did restart again but sometimes it happens. I see you already had them running for some days, one of it could've been almost complete (if you would've crunched it with half cores)! Too bad...
As I did run my tasks recently I suddenly had a bluescreen and had to restart, but it didn't affect the SoBs, they did continue with no problem...

If I was you, I'd dedicate that machine to either do TRP-SV, in which it works like a true 8 core, or cunch the smaller LLR tests such as PPS, SR5 or TRP.

Well, even if AMD works bad on the LLRs it's doable, so people shouldn't be discouraged just because of that. Even turtles reach their goal. ;-)
____________
Life is Science, and Science rules. To the universe and beyond
Proud member of BOINC@Heidelberg

numbermaniac
Volunteer tester
Send message
Joined: 28 Mar 14
Posts: 197
ID: 305955
Credit: 13,067,196
RAC: 11,098
321 LLR Silver: Earned 100,000 credits (225,812)PPS LLR Ruby: Earned 2,000,000 credits (2,116,107)SR5 LLR Amethyst: Earned 1,000,000 credits (1,038,057)SGS LLR Ruby: Earned 2,000,000 credits (3,422,268)TRP LLR Gold: Earned 500,000 credits (502,390)321 Sieve (suspended) Gold: Earned 500,000 credits (512,303)Generalized Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (503,510)PPS Sieve Ruby: Earned 2,000,000 credits (2,002,374)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Bronze: Earned 10,000 credits (34,269)TRP Sieve (suspended) Bronze: Earned 10,000 credits (32,679)AP 26/27 Gold: Earned 500,000 credits (513,461)GFN Ruby: Earned 2,000,000 credits (2,010,259)PSA Silver: Earned 100,000 credits (141,661)
Message 99312 - Posted: 1 Oct 2016 | 0:58:01 UTC

Before shutting down the computer, suspend BOINC if SoB (or any LLR project) is running, because it then sends a signal to the LLR task to checkpoint and stop it right there. Make sure you have the "Leave non-GPU tasks in memory while suspended" OFF, else it won't work and may either crash or start from the beginning when you reboot the computer.
____________
1 PPSE (+2 DC) & 5 SGS primes

Message boards : Number crunching : Extremely frustrated-WU's keep restarting

[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2022 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.48, 0.53, 0.49
Generated 18 Aug 2022 | 10:02:34 UTC