![]() |
|
|
#1 |
|
Jan 2003
7 Posts |
The iteration times for my machine (when left idle) are 0.197 at the lowest. That corresponds to an Athlon XP 1800+ approximately. I'm running a 2400+! what's going on?
My system specs are as follows: Athlon XP 2400+ (2000MHz) - 266MHz FSB 512MB PC2100 ECS K7S5A (SiS 735 chipset) not very relevant: GeForce FX 5200 128MB Maxtor 80GB 7200rpm onboard sound All the latest drivers |
|
|
|
|
|
#2 |
|
Sep 2002
12268 Posts |
Things that can slow prime95 down ( if running Windows )
Video card resolution and color depth. A screen saver. Another process running in idle contending for the cycles. ( Another distributed project ) AntiVirus software ( more than 1 at a time has a dramatic effect ) Some legitimate service or other background process. A virus/trojan/worm/spyware. ActiveX applets running in a browser. The version of Windows 98,ME,2000,XP Very large number of processes running so task switching eating alot of CPU cycles. Time slice too small, not much done before switching so high CPU overhead. WinModem, CPU has to do all calcs that a dedicated chip would do. CPU heavy driver. Not enough memory ( OS has to use HD for virtual RAM ). |
|
|
|
|
|
#3 |
|
Sep 2002
12268 Posts |
Forgot these.
Version of prime95, some are much slower/faster. The size of the Mersenne number working with. Number of banks of memory used. Single, pair, 2 pairs, are they all matched ?, using unmatched in any way ( speed, size ) has a big effect. |
|
|
|
|
|
#4 |
|
Jan 2003
7 Posts |
Thanks for your response!
I've got most of those factors minimized. The tests I've done are with everything pretty much idle. I disable as much as I can before going to school each day and take a look at the iteration times when I return.. They're what I would expect from a maching about 400 to 500MHz lower. However, I have no idea what those benchmark iteration times are based on.. WinXP? version of Prime 95? so.... I won't complain :) |
|
|
|
|
|
#5 |
|
"GIMFS"
Sep 2002
Oeiras, Portugal
2·7·113 Posts |
If you´re referring to the Benchmarks shown in GIMPS "Benchmarks" page, those are based upon version 21 of the program. This means they are worse than what is currently expected, with more efficient versions around. So, the timings you obtained are in fact not corresponding to a XP1800+, but to something even slower.
That said, you may have processes you are not aware of running in the background, and stealing CPU cycles from P95. Try installing TaskInfo or some similar utility, that allows you to keep track of every process running. Your PC may have been victim of some malicious code, sent via email or screen saver, that has installed a program that consumes machine resources. That happened to me once, and mind you, I had that latest version of the antivirus installed! It was a screen saver that had the code. The PC was very slow, unstable, and it wasn´t until I found that odd process running that was able to solve the problem. Just my $0.02... |
|
|
|
|
|
#6 |
|
Apr 2003
California
9210 Posts |
I agree that video settings make a difference. Use the lowest refresh rate, resolution, and color depth that you can stand. See if that helps!
|
|
|
|
|
|
#7 |
|
Sep 2003
Borg HQ, Delta Quadrant
70210 Posts |
Have you tried upping the priority of Prime95? By default it runs at 1 which is the lowest. I usually set mine to 8 since 9 is the default for most apps. But that will put it safely above the screensaver and maybe some other stuff that might steal cycles.
|
|
|
|
|
|
#8 |
|
Apr 2003
California
9210 Posts |
Please post your version of Prime95.
What exponent are you working on? Also see these threads: Software > Slow iteration times with 23.7 Software > System tweaks to speed GIMPS Hardware > Motherboard replaced - now goes slower Hardware > What's your per iteration time? Hardware > computer seems to LAg when Prime95 is running Animated GIFs can also slow you down a lot. However I imagine you've closed down other apps. |
|
|
|
|
|
#9 |
|
Oct 2002
Lost in the hills of Iowa
26×7 Posts |
I pull about .175 with one of my XP1800+ - but it's overclocked to appx. 1800 Mhz (XP2200+ range) at 175ish on the FSB, running PC2700 for RAM and nothing else running on the machine at those timings.
I suspect your XP2400 is hampered a little by the SIS chipset (they're not known for being the most efficient, vs. NForce and most VIA solutions) and you might have more services running than you need, or something running in the background, or an old Prime version, or a combination of the above. |
|
|
|
|
|
#10 |
|
Jan 2003
7 Posts |
I'm running version 23.7.1
Exponent 33512231 Lowest iteration time of 0.197 I'm looking into spyware, program priority, etc. I just haven't had much time lately. Thanks for all the suggestions. |
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Iteration times in i5 and i7 | Jud McCranie | Information & Answers | 53 | 2013-08-17 19:09 |
| What are your per-iteration times? | LiquidNitrogen | Hardware | 22 | 2011-07-12 23:15 |
| Iteration times slow compared to benchmark | cd1940 | Software | 4 | 2004-07-05 17:55 |
| LLR.exe FFT crossovers and iteration times | SlashDude | 15k Search | 0 | 2004-01-28 05:47 |
| Slow iteration times with 23.7 | smoffat | Software | 13 | 2003-10-22 22:50 |