mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Software

Reply
 
Thread Tools
Old 2011-02-23, 01:08   #1
ixfd64
Bemusing Prompter
 
ixfd64's Avatar
 
"Danny"
Dec 2002
California

1001010011102 Posts
Default compiler/assembler optimizations possible?

It seems that a stable release of Prime95 26.x may become available soon. Notably, version 26.x is going to feature (among other things) optimizations for various CPU architectures.

However, some people report getting faster iteration times simply by changing the way Prime95 is compiled. See this post, for example.

Does Prime95 26.x use these optimizations? If not, then an additional 5-10% increase in performance would be quite nice!
ixfd64 is offline   Reply With Quote
Old 2011-02-23, 01:46   #2
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

164758 Posts
Default

Quote:
Originally Posted by ixfd64 View Post
However, some people report getting faster iteration times simply by changing the way Prime95 is compiled.
I put little stock in such reports. All the CPU intensive code is in assembly language. Recompiling C code will not affect timings.
Prime95 is offline   Reply With Quote
Old 2011-02-23, 04:15   #3
CRGreathouse
 
CRGreathouse's Avatar
 
Aug 2006

135418 Posts
Default

Quote:
Originally Posted by Prime95 View Post
I put little stock in such reports. All the CPU intensive code is in assembly language. Recompiling C code will not affect timings.
Or, more precisely: the amount of time saved by recompiling will be smaller than that spent recompiling. :)
CRGreathouse is offline   Reply With Quote
Old 2011-02-23, 08:16   #4
tichy
 
Nov 2010

22×19 Posts
Default

Quote:
Originally Posted by ixfd64 View Post
See this post, for example.
Yes, these were bogus effects. For some reason deviation if timings was more frequently on favour of recompiled binary than the official one.

However, back in december I started tackling the assembler part of mprime. When I reach something more interesting than 1-tick speedup of 32-bit =<74bit factoring code I will share it here. Right now my priority is porting as much as I can to NASM. GAS already proved a viable alternative to MASM, but the chores of adding debugging symbols (I used STABS directives) was to much troublesome.
tichy is offline   Reply With Quote
Old 2011-02-23, 11:30   #5
tichy
 
Nov 2010

22·19 Posts
Default

Quote:
Originally Posted by tichy View Post
bogus
This is a wrong word - not being a native english speaker leads sometimes to such mistakes. I needed an adjective which means that the observed timings improvements were not intrinsicaly related to the code itself but were merely result of the environment the code was ran in.
tichy is offline   Reply With Quote
Old 2011-02-23, 21:43   #6
axn
 
axn's Avatar
 
Jun 2003

115408 Posts
Default

Quote:
Originally Posted by tichy View Post
This is a wrong word - not being a native english speaker leads sometimes to such mistakes. I needed an adjective which means that the observed timings improvements were not intrinsicaly related to the code itself but were merely result of the environment the code was ran in.
In the context that you used, it fits perfectly.
axn is offline   Reply With Quote
Old 2011-02-25, 03:59   #7
cheesehead
 
cheesehead's Avatar
 
"Richard B. Woods"
Aug 2002
Wisconsin USA

22·3·641 Posts
Default

Quote:
Originally Posted by tichy View Post
This is a wrong word - not being a native english speaker leads sometimes to such mistakes. I needed an adjective which means that the observed timings improvements were not intrinsicaly related to the code itself but were merely result of the environment the code was ran in.
It wasn't deliberately bogus, which would have implied deception. It was unintentionally, accidentally (and unknowingly, at that time) bogus -- and most readers would understand that.

Last fiddled with by cheesehead on 2011-02-25 at 04:00
cheesehead is offline   Reply With Quote
Old 2011-02-25, 20:05   #8
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

133568 Posts
Default

Quote:
Originally Posted by cheesehead View Post
and most readers would understand that.
Maybe not some non-native english speakers.
henryzz is online now   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
ARM builds and SIMD-assembler prospects ewmayer Mlucas 183 2019-02-25 08:17
Glibc 2.27 Is Being Released Soon With Numerous Performance Optimizations heliosh Lounge 7 2018-02-02 18:19
Turn off GCC sse-using optimizations? ewmayer Programming 3 2016-09-30 07:15
Msieve 64 bit optimizations.. Carlo Msieve 1 2011-09-09 20:44
northwood optimizations E_tron Software 8 2006-01-08 15:15

All times are UTC. The time now is 17:26.

Sun May 9 17:26:52 UTC 2021 up 31 days, 12:07, 1 user, load averages: 4.29, 4.27, 4.11

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.