𝕏

Optimized app ?

Message boards : Number crunching : Optimized app ?
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next

AuthorMessage
Profile Skivelitis2
Avatar

Send message
Joined: 26 Apr 15
Posts: 4
Credit: 1,550,503
RAC: 17
Message 556 - Posted: 12 Oct 2015, 19:16:29 UTC

Kudos to sesef!! I crunched this project to the 3k level, took 4 days with all 5 of my CPU cores. The next step to the 30k and 42k levels seemed too daunting to me as it would have taken months. Since installing the optimized app, I should reach 42k tomorrow (or when pendings validate) after only 5 days of part-time crunching. After a short break, I may just go after the 300k badge.
ID: 556 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile correcaminos

Send message
Joined: 14 Apr 15
Posts: 2
Credit: 1,051,753
RAC: 0
Message 562 - Posted: 13 Oct 2015, 20:08:41 UTC

Hello

I have a i7 970 ( 1366 ) with 64 W7 .

I downloaded the SSE2 , SSE3 and SSE41 files but I have no where to put files or that I have to do with them.

I need help for noobs

Thank you
ID: 562 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
mikey
Avatar

Send message
Joined: 5 Jul 15
Posts: 13
Credit: 4,908,035
RAC: 0
Message 565 - Posted: 13 Oct 2015, 22:29:25 UTC - in response to Message 562.  

Hello

I have a i7 970 ( 1366 ) with 64 W7 .

I downloaded the SSE2 , SSE3 and SSE41 files but I have no where to put files or that I have to do with them.

I need help for noobs

Thank you


Download the file and extract it into the C:\program data\Boinc\projects\Denis directory and restart Boinc.
ID: 565 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile correcaminos

Send message
Joined: 14 Apr 15
Posts: 2
Credit: 1,051,753
RAC: 0
Message 566 - Posted: 14 Oct 2015, 6:29:25 UTC - in response to Message 565.  
Last modified: 14 Oct 2015, 6:29:48 UTC

Ok finally doing good.

Maximum time less than 5 min

Thanks
ID: 566 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
sesef

Send message
Joined: 22 Apr 15
Posts: 4
Credit: 17,166,398
RAC: 0
Message 581 - Posted: 25 Oct 2015, 11:48:20 UTC
Last modified: 25 Oct 2015, 11:48:33 UTC

AVX2 version.

Download:
http://optos.sesef.pl/denis or https://dl.dropboxusercontent.com/u/1452459/denis/denis1.6.1_avx2.zip

You should get up to ~30% speedup depends on cpu type, on new Intel Skylake even more.
ID: 581 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
mikey
Avatar

Send message
Joined: 5 Jul 15
Posts: 13
Credit: 4,908,035
RAC: 0
Message 583 - Posted: 25 Oct 2015, 23:57:48 UTC - in response to Message 581.  

AVX2 version.

Download:
http://optos.sesef.pl/denis or https://dl.dropboxusercontent.com/u/1452459/denis/denis1.6.1_avx2.zip

You should get up to ~30% speedup depends on cpu type, on new Intel Skylake even more.


My laptop trashed all the units using this one, I reverted to the old one and it is fine again.
ID: 583 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Cartoonman

Send message
Joined: 22 Oct 15
Posts: 3
Credit: 719,262
RAC: 0
Message 584 - Posted: 26 Oct 2015, 2:32:13 UTC - in response to Message 583.  

^ Same result. On W7 i5 2400
ID: 584 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Ross*

Send message
Joined: 8 Jun 15
Posts: 1
Credit: 8,103,612
RAC: 0
Message 585 - Posted: 26 Oct 2015, 5:27:55 UTC - in response to Message 581.  

Hi
They go fine on 5930 +5960s
Ross*
ID: 585 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
sesef

Send message
Joined: 22 Apr 15
Posts: 4
Credit: 17,166,398
RAC: 0
Message 586 - Posted: 26 Oct 2015, 6:31:42 UTC - in response to Message 583.  

AVX2 version.

Download:
http://optos.sesef.pl/denis or https://dl.dropboxusercontent.com/u/1452459/denis/denis1.6.1_avx2.zip

You should get up to ~30% speedup depends on cpu type, on new Intel Skylake even more.


My laptop trashed all the units using this one, I reverted to the old one and it is fine again.


^ Same result. On W7 i5 2400



AVX2 is available only on i3/i5/i7/xeon 4th gen or newer processors. So only i3/i5/i7 4xxx+ or Xeon v3+
ID: 586 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Chilean
Avatar

Send message
Joined: 9 Apr 15
Posts: 11
Credit: 3,149,460
RAC: 0
Message 587 - Posted: 26 Oct 2015, 12:04:11 UTC - in response to Message 586.  
Last modified: 26 Oct 2015, 12:06:23 UTC

AVX2 version.

Download:
http://optos.sesef.pl/denis or https://dl.dropboxusercontent.com/u/1452459/denis/denis1.6.1_avx2.zip

You should get up to ~30% speedup depends on cpu type, on new Intel Skylake even more.


My laptop trashed all the units using this one, I reverted to the old one and it is fine again.


^ Same result. On W7 i5 2400



AVX2 is available only on i3/i5/i7/xeon 4th gen or newer processors. So only i3/i5/i7 4xxx+ or Xeon v3+


Is an AVX "1" version worth it at all? I have a 3rd generation CPU on my main laptop which only support AVX. Not AVX2.
I'll try this app on my 4th generation laptop tho.
ID: 587 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Jim1348

Send message
Joined: 28 Apr 15
Posts: 29
Credit: 1,426,883
RAC: 0
Message 590 - Posted: 26 Oct 2015, 17:13:19 UTC
Last modified: 26 Oct 2015, 17:18:58 UTC

The AVX2 version looks very good to me. For the 600 series, I am getting about 2 minutes 32 seconds running on four cores of an i7-4771 (another core supports a GPU on Folding, and the other three cores are largely free). This is on Win7 64-bit. The core temps are a little higher, which is usual for AVX2 work, averaging about 70 C now.
ID: 590 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Nosferatu*

Send message
Joined: 10 May 15
Posts: 1
Credit: 8,898,939
RAC: 0
Message 607 - Posted: 30 Oct 2015, 9:28:27 UTC

Is there a Windows 32 bit app available? I have 3 machines that are unable to run 64 bit Windows.
ID: 607 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Chilean
Avatar

Send message
Joined: 9 Apr 15
Posts: 11
Credit: 3,149,460
RAC: 0
Message 609 - Posted: 30 Oct 2015, 17:29:31 UTC - in response to Message 607.  

Is there a Windows 32 bit app available? I have 3 machines that are unable to run 64 bit Windows.


Or AVX "1" ? :)
ID: 609 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
[AF>FAH-Addict.net]toTOW

Send message
Joined: 11 Apr 15
Posts: 24
Credit: 4,161,347
RAC: 576
Message 610 - Posted: 31 Oct 2015, 15:07:15 UTC

I'm not convinced by the AVX2 application ... on my i7 4710HQ, it's actually slower than the previous application :(

The only case where it's faster is if I run it on only 4 cores instead of 8 threads. But in this case, the speed improvement is not enough to compensate the loss of the 4 other processes. Doing 4 WUs every 9min30 is finally producing less than doing 8 WUs in 12 minutes (with previous application, it varies between 12 and 15 minutes with the new one).
ID: 610 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
kain

Send message
Joined: 16 Apr 15
Posts: 20
Credit: 4,958,309
RAC: 7,848
Message 611 - Posted: 31 Oct 2015, 15:10:24 UTC - in response to Message 610.  

I'm not convinced by the AVX2 application ... on my i7 4710HQ, it's actually slower than the previous application :(

The only case where it's faster is if I run it on only 4 cores instead of 8 threads. But in this case, the speed improvement is not enough to compensate the loss of the 4 other processes. Doing 4 WUs every 9min30 is finally producing less than doing 8 WUs in 12 minutes (with previous application, it varies between 12 and 15 minutes with the new one).


It's normal behaviour.
ID: 611 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
[AF>FAH-Addict.net]toTOW

Send message
Joined: 11 Apr 15
Posts: 24
Credit: 4,161,347
RAC: 576
Message 613 - Posted: 31 Oct 2015, 16:21:58 UTC

So this kind of optimizations shouldn't be used on HT processors ?

It might be complicated to include such logic in assignement process or in code logic :(
ID: 613 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Curious

Send message
Joined: 18 Oct 15
Posts: 3
Credit: 210,007
RAC: 0
Message 615 - Posted: 1 Nov 2015, 1:54:15 UTC - in response to Message 613.  
Last modified: 1 Nov 2015, 2:04:08 UTC

Yes, in Asteroids@Home message boards they briefly explain why AVX isn't suggested for CPU with Hyper Threading, see here. Moreover on Primegrid message boards is pointed out that simulating a duble number of cores causes the chip to produce more heat. I've noticed that AVX2 is significantly slower than SSE4.1 on my HT Intel (Haswell) CPU too.

EDIT
URL added
ID: 615 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Jim1348

Send message
Joined: 28 Apr 15
Posts: 29
Credit: 1,426,883
RAC: 0
Message 619 - Posted: 1 Nov 2015, 15:52:59 UTC
Last modified: 1 Nov 2015, 15:58:42 UTC

My tests show the following on an i7-4771 CPU, Win7 64-bit:
Comparison of Sesef's DENIS optimizations on the "3XP 1800" series work units:

With Sesef 1.6.1 AVX2 optimization:
DENIS running on 8 virtual cores - 9 minutes 42 seconds (CPU temp - 63 C average)
DENIS running on 4 virtual cores (other 4 cores free) - 7 minutes 7 seconds (CPU temp - 55 C average)

With Sesef 1.5.5 SSE3 optimization:
DENIS running on 8 virtual cores - 10 minutes 43 seconds (CPU temp - 60 C average)
DENIS running on 4 virtual cores (other 4 cores free) - 8 minutes 9 seconds (CPU temp - 54 C average)

So in each case, the AVX2 optimization is faster than the SSE3 optimization. I doubt that Sesef would have released it otherwise.
However, the temps can build up, especially if you have a GPU card. That could cause throttling of the CPU in some cases, thus lowering its speed.
ID: 619 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Curious

Send message
Joined: 18 Oct 15
Posts: 3
Credit: 210,007
RAC: 0
Message 620 - Posted: 1 Nov 2015, 17:22:28 UTC - in response to Message 619.  
Last modified: 1 Nov 2015, 17:26:00 UTC

I assume you don't know about Crunch3r's SSE4.1 app version. This is the one I'm referring to (not sesef's SSE3 one): it's two times faster than AVX2 on my CPU when running one WU at a time without any other application (distributed computing nor not-DC) so no throttling at all. I know it's a really simple scenario but it allows you to understand things easily.
Sure enough on other CPUs it will perform differently than my Haswll CPU with Hyper Threading and factory power limitation (which I removed through Intel extreme tuning utility though), but I wrote it clearly that I was referring to my particular case.

EDIT
Corrected typos
ID: 620 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Trotador

Send message
Joined: 9 Apr 15
Posts: 11
Credit: 79,474,142
RAC: 18,238
Message 621 - Posted: 1 Nov 2015, 17:47:49 UTC

I would like to test AVX2 application but it seems to be windows only.

Is it possible to make a linux version?
ID: 621 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next

Message boards : Number crunching : Optimized app ?