![]() |
|
|
#23 | |
|
Aug 2002
3110 Posts |
Quote:
When I tried to boost the memory speed (256M DDR 2100) on my KG7-R, Athlon 1400, memtest86 was very happy for all memory modes of the mobo (turbo, etc...). Basically, it changes the latency, the driver strength, etc... However, as soon as I tried to boot windows, it crashed each time until I put the memory back to normal mode. Windows could boot more or less depending on the speed setup of the memory. So, I am skeptical about relying on memtest86 only. Unless I missed something. I assume memtest86 can bypass the cache, etc..., doesn't it ? I hope the tests were performed on the external memory, not in the cache... |
|
|
|
|
|
|
#24 |
|
Sep 2002
2×7 Posts |
Lord_Humungus
outlnder Lycorn Deamiter DSheets daWabbit Thank you one and all. When I returned home from work this morning Prime95 was still wizzing along at 0.105 sec per iteration, and my other programs responded as they should, shutdown now taking seconds rather than minutes. I don't understand why everything but Prime95 slowed to a crawl on starting up with my new DDRAM or how long it lasted, but I am glad to be back to normal after about two weeks of frustration, To get back on topic my CPU temp is 58C with Prime95 running, and 53C when idle, Motherboard temp 30C constant. John |
|
|
|
|
|
#25 |
|
Aug 2002
2568 Posts |
I suspect you totally missed everything.
Unless you have a very large L2 cache, eg, 128MB+ I am pretty sure memtest runs on external ram. ALso you get to select how much memory you test also. By the way dont ever settle for 100% memory coverage, you need at LEAST 4000%-8000% for short terms, and most likely longer if you want complete stability. Just run the window client and not the DOS one. Your problem solved. If you cant even get into windows I dont see why you are trying to test. |
|
|
|
|
|
#26 |
|
Aug 2002
111112 Posts |
My mistake. I was referring to an other tool that doesn't need DOS to test the memory. It wasn't memtest.
The thing is that it seemed to work fine but windows couldn't boot afterwards. So, it didn't help me much. Thanks for your help. |
|
|
|
|
|
#27 |
|
Oct 2002
Lost in the hills of Iowa
26·7 Posts |
Athlon XP1800+ oc to 1610 (140FSB) - Alpha 8045 / Delta 68CFM
Case has another Delta 68, the stock PS fan, and a 35-ish CFM Sunon blowing air through it - chipset temp is usually 1 or 2 degrees C above ambient, CPU 13 or 14 above chipset. Which means on warm days the CPU sees 55-ish - I do NOT have A/C in my computer room, thus the serious cooling. I haven't gotten around to setting up LMmonitor on any of the Linux boxen - yet - but I don't overclock the s2460 and most of the rest are PPGA Celerons, K5s, or K6s - and the cooling on all of THOSE machines is definite overkill - hand-me-down Athlon-class cooling on the Celerons and K6s, and hand-me-down K6-class cooling on the K5s. I've not had any errors on the CPUs/MBs I could attribute to heat - couple dead hard drives, but those were old and might have just been age deaths. *ALL* of my machines have been RC5/OGR/CSC/DES crunchers 'till a few months back, when I swapped a couple of them over to UD. Got fed up with UD not answering questions and generally not being responsive to their volunteers, so now I'm moving machines over here. Well, most of them - I still have one WFWg 3.11/DOS machine for some of my *old* won't run on 95/98 games - it's probably gonna stick with d.net 'till it finally dies. 9-) |
|
|
|
|
|
#28 |
|
"GIMFS"
Sep 2002
Oeiras, Portugal
158210 Posts |
QuintLeo
Be very welcome to our GIMPS community! Hope you will enjoy "making math history" as much as I have been enjoying over the last few months (joined GIMPS in January). |
|
|
|
|
|
#29 | |
|
Aug 2002
2·5 Posts |
Quote:
|
|
|
|
|
|
|
#30 |
|
Aug 2002
23·32 Posts |
That motherboard uses the sis735 chipset, and i have 20 of them running various processors and all on prime95. it is really stable, with bioses dated after or on dec 10 2001. The bios of 9/18(?) gives errors on anything faster than an 1800+.
|
|
|
|
|
|
#31 |
|
Sep 2002
2·7 Posts |
Recently I began to get errors with other programs when running Prime95, the same faults that I previously cured by changing from SDRAM to DDRAM!
I updated the BIOS to the latest version as recommended giving a slight improvement. Mine is a sorry tale, when I got around to checking the cpu temp it was 65C with prime running droping to the low 50's without. I opened the box and found a build up of dust on the stock HS, off to the store for a big copper HS and high speed fan and case fan, the dust is travelling too fast to stick! CPU temp now between 45C and 50C depending on ambient. My problem now is the noise, my family are threatening to buy me a shed for me and my box, I won't have any cooling problems outside! Sound familiar? John |
|
|
|
|
|
#32 |
|
Aug 2002
378 Posts |
I had a similar problem earlier this year. Temperature starting going up abnormally. Cause: dust build-up. Now, I also have my case opened to keep it cooler and it builds up more dust everywhere so I have to inspect the motherboard once in a while and clean it with a spray and a vacuum cleaner. Talk about a work!
As for the noise and family issues... welcome to the club. Try to put your PC in a location where they don't see it too much, otherwise, you are going to have problems! :D |
|
|
|
|
|
#33 |
|
Aug 2002
2×3×53 Posts |
Imagine what my apartment sounds like with 17 computers running 24/7 in the dining room.
The only real choice you have is to get a quieter CPU fan. Most likely you got a copper heatsink with something like a 50mm Delta 48CFM fan. The Delta is putting out close to 48 db of high pitched whine. I replaced my loud CPU HSFs with a Volcano 9 setup. Big heatsink with copper insert and an 80mm fan. It is adjustable so you can get a quieter room. Not trying to promote anything, but there are quite a number of Forums that deal with this problem. Check into http://arstechnica.infopop.net/OpenT...amp;f=77909585 for some terrific info. |
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Interesting GPU heat issue... | nucleon | GPU Computing | 20 | 2012-01-29 19:25 |
| Four heat seeking missiles | mfgoode | Puzzles | 51 | 2007-01-10 10:02 |
| Heat problems | michaf | Sierpinski/Riesel Base 5 | 3 | 2006-07-04 21:11 |
| Opertons x52, x54 (3000 CPUs) heat = floating point errors | dsouza123 | Hardware | 5 | 2006-05-01 17:05 |
| Keeping the Heat down at Home | outlnder | Hardware | 61 | 2003-02-15 03:12 |