![]() |
|
|
#1 |
|
Sep 2002
Austin, TX
3·11·17 Posts |
are there any further pentium 4 northwood optimizations between versions 23.4 and the current version, 24.14?
Im wondering if my northwoods should be updated. Most are not HT enabled. |
|
|
|
|
|
#2 |
|
P90 years forever!
Aug 2002
Yeehaw, FL
11101011001102 Posts |
Yes. Version 23 was written for a 256K L2 cache. Version 24 makes use of the Northwood's larger L2 cache for many FFT lengths.
|
|
|
|
|
|
#3 |
|
Aug 2002
Termonfeckin, IE
22×691 Posts |
Yes definitely update to v24 for all Pentium4s and Athlon XPs and above.
|
|
|
|
|
|
#4 |
|
"Patrik Johansson"
Aug 2002
Uppsala, Sweden
52×17 Posts |
I just upgraded my clients at home to 24.14 and got good improvements for many computers. However, my Celeron D went from 0.034 s to 0.042 s per iteration for my double checks, so I went back to version 23.9 which I was running earlier. Which version is best to run for a Celeron D?
|
|
|
|
|
|
#5 |
|
"Patrik Johansson"
Aug 2002
Uppsala, Sweden
52·17 Posts |
OK, I just saw that there was only one more recent version to download, v 24.13, and that one also turned out to be slower. I think I'll have to stick with version 23.9.
For 896K FFT length on a Celeron D 3.06@3.335 GHz I get: Code:
Version Iteration time 23.9 0.034 s 24.13 0.040 s 24.14 0.042 s |
|
|
|
|
|
#6 |
|
P90 years forever!
Aug 2002
Yeehaw, FL
2·53·71 Posts |
Is your Celeron D a 128K L2 cache?
|
|
|
|
|
|
#7 |
|
Feb 2005
Colorado
22×7×23 Posts |
I just did my own test on a Celeron D running at 3.57 Mhz:
For a 896K FFT: V23.9: 28.254 ms. V24.14: 25.992 ms. This benchmark was done with the static linked mprime, not Prime95. George, all Celeron D's are 256K cache. Also, the fastest 128K Celeron was 2.8 Ghz. |
|
|
|
|
|
#8 | |
|
P90 years forever!
Aug 2002
Yeehaw, FL
752610 Posts |
Quote:
Last fiddled with by Prime95 on 2006-01-07 at 16:36 |
|
|
|
|
|
|
#9 |
|
"Patrik Johansson"
Aug 2002
Uppsala, Sweden
1101010012 Posts |
The times dropped after reboot, so now 24.14 is faster. Thanks for the info.
Code:
Version Iteration time 23.9 0.032 s 24.14 0.030 s Last fiddled with by patrik on 2006-01-08 at 15:16 |
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| 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 |
| compiler/assembler optimizations possible? | ixfd64 | Software | 7 | 2011-02-25 20:05 |
| Northwood, Prescott Or Extreme Edition? | georgekh | Hardware | 13 | 2005-03-17 06:31 |
| Northwood benchmarks... | Xyzzy | Hardware | 7 | 2003-03-25 00:36 |