![]() |
|
|
#23 | |
|
Apr 2003
Berlin, Germany
36110 Posts |
Quote:
In 2D benchmarks the dual channel gives some 0.x to 2.x % performance increase over single channel - as expected. 3D stuff is much more bandwidth hungry but the 100-200 MB/s for average 2D display modes have little impact. But they are still too much for PC133 while being only a little overhead for 2xDDR2700. And the Athlon FSB can't even fetch that much (2.7 or 3.2GB/s in the latest models). Matthias |
|
|
|
|
|
|
#24 | |
|
Oct 2002
Lost in the hills of Iowa
26·7 Posts |
Quote:
I think it may depend on the video card, though - *thought* I was running 1600x1280, but turns out I'm running 1600x1200, but I distinctly remember it being 1600x1280 on the first machine I used that was capable of higher than 1280x1024 resolution. <mock glares at NVidea, for depriving me of my rightfull extra 80 lines of resolution> 9-) |
|
|
|
|
|
|
#25 | |
|
"Patrik Johansson"
Aug 2002
Uppsala, Sweden
52×17 Posts |
Quote:
|
|
|
|
|
|
|
#26 |
|
Sep 2002
Vienna, Austria
3·73 Posts |
My Per-iteration time reduced to 49ms With Prime95 V23.4.......... :(
Hopes that George will do further improvements on the P4 code. |
|
|
|
|
|
#27 | |
|
"GIMFS"
Sep 2002
Oeiras, Portugal
2·7·113 Posts |
wpolly wrote:
Quote:
|
|
|
|
|
|
|
#28 |
|
Apr 2003
California
22×23 Posts |
If you are using onboard video, the video settings can make a big difference, like 25%!
Work-around: right-click an empty area of the desktop, choose Properties, Settings tab. Use minimal video settings like 640*480, 8-bit color. On the Advanced button, go to the monitor tab and choose the lowest refresh rate. Fix: get a PCI or AGP video card. See the thread http://www.mersenneforum.org/viewtopic.php?t=830 . Edit: sometimes 8-bit color may not give you the best performance, if the CPU has to dither 16 bits down to 8 bits. It's best to check a few settings for your best performance. |
|
|
|
|
|
#29 |
|
Sep 2002
Vienna, Austria
3×73 Posts |
With 1024*768: 51.220ms
With 800*600: 54.765ms Looks like that The performance INCREASES with the resolution!!!??? PS: My computer is P4 2.0G/512M/GF4 MX440, working on 1024K FFT Size. |
|
|
|
|
|
#30 | |
|
Banned
"Luigi"
Aug 2002
Team Italia
10011000000012 Posts |
Quote:
Luigi |
|
|
|
|
|
|
#31 | |
|
Aug 2003
32 Posts |
Quote:
EDIT: moved my iteration times to the benchmark thread BP 0007 |
|
|
|
|
|
|
#32 |
|
Aug 2002
2×32×13×37 Posts |
The mobile P4-based Celerons have 256KB also...
|
|
|
|
|
|
#33 | |
|
Aug 2003
Upstate NY, USA
2×163 Posts |
Quote:
640x480 800x600 848x480 1024x768 1152x864 1280x720 1280x768 1280x960 1280x1024 1360x768 1600x900 1600x1024 1600x1200 Also, here are some time trials for 10 iterations of 50,000,000 on my P4 2.4 Resolution - Colors - Time (s) 1152x864 - 32bit - 1.784 1152x864 - 16bit - 1.775 1152x864 - 256 - 1.782 1024x768 - 32bit - 1.798 - my regular setting 1024x768 - 16bit - 1.778 1024x768 - 256 - 1.803 800x600 - 32bit - 1.777 800x600 - 16bit - 1.777 800x600 - 256 - 1.788 So I'm running at 98.7% time efficiency, oh well |
|
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Per iteration time | Jwb52z | PrimeNet | 6 | 2011-09-09 04:06 |
| iteration time under XP | Unregistered | Software | 20 | 2004-09-30 06:35 |
| WHAT IS PER ITERATION TIME? | MavsFan | Software | 1 | 2003-12-12 02:35 |
| iteration time log | crash893 | Software | 1 | 2002-11-13 05:45 |
| Per iteration time | sofII | Software | 8 | 2002-09-07 01:51 |