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
1) Message boards : Problems and Help : Want to remove all CPU work only. (Message 134337)
Posted 327 days ago by Sandman192Project donor
I try to remove all CPU work and leave GPU for Prime. But every time I get 321 Sieve (321-Sieve) check right back up. I can't uncheck it?
2) Message boards : Problems and Help : Prime is ignoring 50% CPU usage. (Message 133781)
Posted 345 days ago by Sandman192Project donor
Ok, forget the app_config.xml config setting since it's set for a different LLR anyway that was for a different Prime challenge long ago.

And running Fermat Divisor Search LLR at 100% CPU and running at all 12 cores with h/p seems to run just fine.

Still, don't know why 50% of CPU is not running at 50%.
3) Message boards : Problems and Help : Prime is ignoring 50% CPU usage. (Message 133723)
Posted 346 days ago by Sandman192Project donor
Ok, why is Prime Fermat Divisor Search LLR (PPS-DIV) ignoring my 50% CPU usage for BOINC Computing Performance? Note: restarting BOINC did not fix it after 50% was changed.
I had it on 50% before it got work and it ran 1 Prime LLR at 6 threads only. I then set it to 100% CPU and it ran 2 Primes LLR at 6 threads. If that's stable to do that then make it do that?
So I tried to use the app_config in Prime to keep it at 6 threads. But it didn't work. So I deleted the app_config and set BOINC Computing Performance to 50% usage because of All the Primes Stayed at 12 threads max regardless of 50% CPU.

I got news that I don't need to add the Prime app_config.xml. But I wish it would run the Divisor Search LLR app_config.xml again when Prime runs into that work again (like it's doing now) so I won't have to change settings in BOINC manager or app_config. Because I heard it would be unstable at 100% on hyperthreading alone w/o overclocking.
BOINC should be Set it and forget it. We shouldn't have to change anything to get what you chang in your work.

Just bring back the app_confing.xml. It worked! Or just put it by default on your Prime website for Fermat Divisor Search LLR (PPS-DIV) and others that you know that would be unstable with hyperthreading and overclocking.
4) Message boards : Problems and Help : Fermat Divisor Search LLR (PPS-DIV) not getting work Oct 10 at 13:46 UTC -6 (Message 133717)
Posted 346 days ago by Sandman192Project donor
I'm not getting any work yet and it's 13:46 UTC -6.
I have an i7 5930 Haswell-E/EP.
I have it set just for Fermat Divisor Search LLR (PPS-DIV) CPU checked box 3 days ago.
I set it for 10 days of work.
CPU at 50%.
The only work I still have is 10 days of GPU work. Right now it's just running 4 Einstein@Home GPU work only. 2 per GPU.

Is my time still off?

Never mind. I had 3 Primes in multi-thread suspend for CPU
Moderator, you can delete this thread if you want.
5) Message boards : Number crunching : World Maths Day Challenge (Message 133715)
Posted 346 days ago by Sandman192Project donor
Sad once more we have 1000's of Tasks the have been aborted.
Maybe someone posting of how to run PG should be posted.
No sense in downloading a few hundred tasks and then turn around and abort them all and then downloading a few more hundred tasks only to abort them as well.
Maybe PG can set aborted tasks to be given back to the same host that has aborted them....

Agreed.
6) Message boards : The Riesel Problem : LLR TRP just stop working (Message 130032)
Posted 478 days ago by Sandman192Project donor
That's still too high. From your times, your CPU is clearly throttling (my older AVX-inferior 6 core SBE outperforms yours, it shouldn't) and unstable. What cooler are you using?

Whether it's air or water/AIO, take it off, clean it out, and put down new grease when you reinstall. Everyone should do a reinstall every 2-3 years. If you can't find any thermal grease ("paste"), buy a tube of Noctua's NT-H2, or NT-H1 if you can't find it. Inexpensive, and very good.


Even at 82c, it's very stable (nothing has crashed) and everything is clean and the past is less than a year old and done by a professional along with the rest of the PC (I will thank about using the past you suggested and maybe in the long run). Tubo overclock from 3.5 to 4Ghz using AI overclocking and it asks what colorer I'm using which is air cooling with all fans on high even the GPUs. It's 80F in my room with AC but I like that temp from time to time.

Note: I am also running 2 GPU work as well.
7) Message boards : The Riesel Problem : LLR TRP just stop working (Message 129950)
Posted 481 days ago by Sandman192Project donor
Turning off HT made my CPU 18c cooler and still running at 100%. From 100c to 82c.
8) Message boards : The Riesel Problem : LLR TRP just stop working (Message 129937)
Posted 481 days ago by Sandman192Project donor
Boinc Benchmarks

Benchmark results:
Number of CPUs: 6 HT off
4980 floating point MIPS (Whetstone) per CPU
18195 integer MIPS (Dhrystone) per CPU

Benchmark results:
Number of CPUs: 12 HT on
4433 floating point MIPS (Whetstone) per CPU
18034 integer MIPS (Dhrystone) per CPU
9) Message boards : The Riesel Problem : LLR TRP just stop working (Message 129917)
Posted 482 days ago by Sandman192Project donor
I had a couple of TRPs that stop working and the CPU is idle and % of work not changing.
So I just aborted them both.

I'm running them on Intel Haswell 12 core total with HT on with multicore work enabled with turbo overclock at 3.99 GHz at 100c package temp.
10) Message boards : Number crunching : NVIDIA driver out v365.10 mabe with fix. (Message 95056)
Posted 1590 days ago by Sandman192Project donor
New version came out v365.19 video drivers. Hope this time it works out.


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2020 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 2.83, 2.53, 2.37
Generated 21 Sep 2020 | 9:48:43 UTC