mersenneforum.org  

Go Back   mersenneforum.org > New To GIMPS? Start Here! > Information & Answers

Reply
 
Thread Tools
Old 2011-07-04, 11:46   #1
Unregistered
 

1ACE16 Posts
Default Support AVX

Support the next version of Prime95 AVX? If yes, is already known when this version is available?
  Reply With Quote
Old 2011-07-04, 13:24   #2
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2×33×139 Posts
Default

Yes to AVX. No as to a known date.

At present, FFT sizes 32, 128, 512, and 2048 are coded and mostly working.
Prime95 is offline   Reply With Quote
Old 2011-07-04, 16:57   #3
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

22·13·113 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Yes to AVX. No as to a known date.

At present, FFT sizes 32, 128, 512, and 2048 are coded and mostly working.
Any hints as to speed increase on these FFTs?
henryzz is offline   Reply With Quote
Old 2011-07-04, 18:02   #4
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2·33·139 Posts
Default

Just got numbers for the 2K FFT:

SSE2: 100 iters in 1.110 ms
AVX: 100 iters in 0.682 ms

I'm not sure if these numbers will extrapolate to large FFTs. These small FFTs operate completely in the L1 cache.
Prime95 is offline   Reply With Quote
Old 2011-07-04, 19:11   #5
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

22·13·113 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Just got numbers for the 2K FFT:

SSE2: 100 iters in 1.110 ms
AVX: 100 iters in 0.682 ms

I'm not sure if these numbers will extrapolate to large FFTs. These small FFTs operate completely in the L1 cache.
Nice. I use the smaller FFTs most of the time in LLR anyway so I am not too worried.
henryzz is offline   Reply With Quote
Old 2011-07-05, 17:12   #6
ixfd64
Bemusing Prompter
 
ixfd64's Avatar
 
"Danny"
Dec 2002
California

239010 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Just got numbers for the 2K FFT:

SSE2: 100 iters in 1.110 ms
AVX: 100 iters in 0.682 ms
These timings seem consistent with the fact that AVX is not twice as fast as SSE2 despite the increase to 256 bits. Hopefully AVX2 and the new Tri-Gate chips will "fix" this.
ixfd64 is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Support Organizations xilman Soap Box 3 2017-04-27 08:13
Crowdfundings we support. chappy Lounge 0 2017-02-18 01:18
AMD Linux support fivemack GPU Computing 1 2015-12-11 03:28
5+ GPU support TheMawn GPU Computing 3 2014-07-13 02:31
Athlon64 support? JuanTutors Software 1 2004-06-04 02:46

All times are UTC. The time now is 10:37.

Sun Jun 20 10:37:08 UTC 2021 up 23 days, 8:24, 0 users, load averages: 3.13, 2.25, 1.79

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.