Prime95 version 27.7 / 27.9 - Page 27

Mark Rose · 2014-02-01, 05:36

Quote:

Originally Posted by Xyzzy

PS - we have also adjusted the variables for mfaktc. Surprisingly, we can run 128/8 in Ubuntu whereas in Windows we had to use 32/8 to get decent screen performance.

128/8 does give me the best performance for my cards (GTX 760 and GTX 430 x 2).

I normally control-c mfaktc when I use my desktop.

I would try setting NumStreams=2 (or even 1) to increase interactive performance at the expense of slightly reduced mfaktc performance.

Windows and Linux do indeed do graphics very differently behind the scenes. The Linux stack has ancient origins, and it's currently being rewritten (Wayland project) but will take a couple more years to be adopted in mainstream distributions.

Xyzzy · 2014-02-01, 15:38

Quote:

For the sake of experimentation, try with HyperThreading turned on. If the source of your lag is competition for CPU resources then HyperThreading should address that very issue.

No change.

Xyzzy · 2014-02-01, 15:45

Quote:

I would try setting NumStreams=2 (or even 1) to increase interactive performance at the expense of slightly reduced mfaktc performance..

We tried values from 1 to 10 without noticing a difference. (This setting is not used when "SieveOnGPU=1", right?)

Xyzzy · 2014-02-01, 16:22

Further testing:

To test lag we did the two following tasks:

1 - Scroll up and down a long web page in Chrome.
2 - Scroll up and down a large PDF file in Document Viewer.

no mfaktc & no mprime = no lag
no mfaktc & mprime (4 cores) = no lag
mfaktc (4/4 = 332-341GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/4 = 332-341GHz-d/d) & no mprime = no lag
mfaktc (4/8 = 375-377GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/8 = 375-377GHz-d/d) & no mprime = no lag
mfaktc (4/16 = 399-400GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/16 = 399-400GHz-d/d) & no mprime = no lag
mfaktc (4/32 = 419-423GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/32 = 419-423GHz-d/d) & no mprime = no lag
mfaktc (4/64 = 435-436GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/64 = 435-436GHz-d/d) & no mprime = no lag
mfaktc (4/128 = 441-442GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/128 = 441-442GHz-d/d) & no mprime = no lag

The above data suggests that we can get no lag if we run neither mfaktc or mprime, or if we just run mprime, or if we just run mfaktc. Once we combine both programs, we get lag.

Mark Rose · 2014-02-01, 19:43

That's odd to me. I always get lag when running mfaktc, regardless of CPU usage, so I control as I said.

One other thing to try is changing the CPU affinity of mprime. By default it binds to core 0, even if running just one process. This is problematic because some interrupts are only handled on core 0, and I'm guessing including interrupts involved with the graphics card.

richs · 2014-02-11, 04:21

This was an interesting error, Windows 64 version 27.7 build 2:

Iteration: 2768128/31272611, ERROR: ROUND OFF (5.219543642e+180) > 0.40
Continuing from last save file.

Now every iteration I get:

Iteration: 17076000 / 31272611 [54.60%]. Per iteration time: 0.031 sec.
Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
Confidence in final result is fair.

What causes such a massive roundoff error? I'm hoping this DC is worth the effort....

Prime95 · 2014-02-11, 04:52

Is this a multi-threaded LL test? If so, there was a bug fixed in 27.9 that displayed massive roundoff errors such as yours. The bug did not affect the final result of the LL test.

LaurV · 2014-02-11, 05:41

Quote:

Originally Posted by richs

Iteration: 2768128/31272611, ERROR: ROUND OFF (5.219543642e+180) > 0.40

Please update to v27.9, or 28.3, there was a bug fixed for the case when you run one worker in many cores. You can read in the current thread, starting with post 71 more details.

edit: whoops! George was faster (I opened the "reply" earlier, I didn't reply/post because of busy-bees here, now is lunch break)

maxxbot · 2014-02-13, 18:33

How do I get multiple worker windows to cooperate on the same exponent? I read that this was possible but cannot figure out how to do it.

Also, is there any way to get prime95 to work on a list of exponents? So far I have had to just enter them in manually one at a time.

richs · 2014-02-22, 19:11

The DC completed successfully. I will update my version. Thanks for the advice!

P.S. - The download software page shows two links for Windows 64-bit software under Upgrade Instructions for Existing Users. The second link should be labelled Windows 32-bit.

Quote:

Originally Posted by richs

This was an interesting error, Windows 64 version 27.7 build 2:

Iteration: 2768128/31272611, ERROR: ROUND OFF (5.219543642e+180) > 0.40
Continuing from last save file.

Now every iteration I get:

Iteration: 17076000 / 31272611 [54.60%]. Per iteration time: 0.031 sec.
Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.
Confidence in final result is fair.

What causes such a massive roundoff error? I'm hoping this DC is worth the effort....

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Prime95 version 27.3	Prime95	Software	148	2012-03-18 19:24
Prime95 version 26.3	Prime95	Software	76	2010-12-11 00:11
Prime95 version 25.5	Prime95	PrimeNet	369	2008-02-26 05:21
Prime95 version 25.4	Prime95	PrimeNet	143	2007-09-24 21:01
When the next prime95 version ?	pacionet	Software	74	2006-12-07 20:30

2014-02-01, 16:22	#290
Xyzzy "Mike" Aug 2002 2·23·179 Posts	Further testing: To test lag we did the two following tasks: 1 - Scroll up and down a long web page in Chrome. 2 - Scroll up and down a large PDF file in Document Viewer. no mfaktc & no mprime = no lag no mfaktc & mprime (4 cores) = no lag mfaktc (4/4 = 332-341GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/4 = 332-341GHz-d/d) & no mprime = no lag mfaktc (4/8 = 375-377GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/8 = 375-377GHz-d/d) & no mprime = no lag mfaktc (4/16 = 399-400GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/16 = 399-400GHz-d/d) & no mprime = no lag mfaktc (4/32 = 419-423GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/32 = 419-423GHz-d/d) & no mprime = no lag mfaktc (4/64 = 435-436GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/64 = 435-436GHz-d/d) & no mprime = no lag mfaktc (4/128 = 441-442GHz-d/d) & mprime (4 cores) = lag; mfaktc (4/128 = 441-442GHz-d/d) & no mprime = no lag The above data suggests that we can get no lag if we run neither mfaktc or mprime, or if we just run mprime, or if we just run mfaktc. Once we combine both programs, we get lag.

2014-02-01, 19:43	#291
Mark Rose "/X\(‘-‘)/X\" Jan 2013 B72₁₆ Posts	That's odd to me. I always get lag when running mfaktc, regardless of CPU usage, so I control as I said. One other thing to try is changing the CPU affinity of mprime. By default it binds to core 0, even if running just one process. This is problematic because some interrupts are only handled on core 0, and I'm guessing including interrupts involved with the graphics card.

2014-02-11, 04:21	#292
richs "Rich" Aug 2002 Benicia, California 2×659 Posts	This was an interesting error, Windows 64 version 27.7 build 2: Iteration: 2768128/31272611, ERROR: ROUND OFF (5.219543642e+180) > 0.40 Continuing from last save file. Now every iteration I get: Iteration: 17076000 / 31272611 [54.60%]. Per iteration time: 0.031 sec. Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4. Confidence in final result is fair. What causes such a massive roundoff error? I'm hoping this DC is worth the effort....

2014-02-11, 04:52	#293
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 7,537 Posts	Is this a multi-threaded LL test? If so, there was a bug fixed in 27.9 that displayed massive roundoff errors such as yours. The bug did not affect the final result of the LL test.

2014-02-13, 18:33	#295
maxxbot Feb 2014 1 Posts	How do I get multiple worker windows to cooperate on the same exponent? I read that this was possible but cannot figure out how to do it. Also, is there any way to get prime95 to work on a list of exponents? So far I have had to just enter them in manually one at a time.