proving that bad ram IS bad ram

[deep_sky]

I have pc100 128 mb ram module that is potentially bad. Win98 detects only 120 mb of it as well. I ran a benchmark on my memory subsystem last night and it sucked. my amd k6 450 with supposedly 128 mb of fast ram was worse than a winchip 266 with 64 mb of ram. How embarrassing! Does anybody know of a place to download a free diagnostic that checks ram for errors? I need to determine that the ram is bad so that I can make my case to the manufacturer of my system (it is two weeks old)

[Mntsnow]

Deepsky,

I bet you have on-board video. and it is using 8megs of "shared" memory. (at the bios screen I bet it shows the memory count correctly) what model of motherboard do you have?

2. "sondra" has a known problem of benchmarking ram with AMD chips. to get the "correct" results you need to multiply the results that sondra gives you by approx. 2x to get "real life" you can get the correct multiplier from the sisoft website.

Mntsnow

[This message has been edited by Mntsnow (edited 10-19-1999).]

[Vampiel]

yeesh no wonder it showed my RAM so low!

[deep_sky]

well if what you guys say is true, then why is my system so slow then. Specifically I am running seti@home and it is performing the calculations much slower than even a slower p2 with half my memory. Other code optimization of the program, I cannot think why my system would be having such a hard time with it.

Yes I do have on-board video and the agp has 8 megs on it. I have not checked what manufacurer it is because the board is so small that they couldnt fit the name on it!!
*sigh* but thanks much for the info. if there is anything else anyone can tell me, i would appreciate it...

[Mntsnow]

Well SKY,
With all your knowledge of the distributed cracking you would also know that the Seti@home cracking client is NOT truly optimized for the AMD chip (another reason I dont do seti /forum/wink.gif) That is why you dont process seti quickly compared to a PII chip. Another reason your system may seem slow is because of your motherboard (AGP graphics ON-board it makes me think you have a PC-Chips motherboard) (they are one of the largest manufacturers of motherboards (makes them under MANY names) These boards are ok but they are FAR from great! you can take your memory/cpu and other componants and put them in a system with a "better" motherboard and you will get "better" benchmarks due to better chipsets and motherboard construction.

If you have any further questions I and others will be happy to enlighten you /forum/biggrin.gif

Mntsnow

[deep_sky]

for your information SNOW
I do happen to know that the code is not optimized for AMD. I know other people who are running the same system and are not bogging down as much.
Well, yes , basically a new system is the answer to my problems, but this a "new system" and so i dont have the money to do so. it will have to happen next year, by which time i will probably ditch the mobo, and get a "real" system =-)
thanks for the info though....

[Mntsnow]

Sky,

Here is some hints that might make Seti work faster for you.
This is from seti's website
What can I do to make SETI@home run faster?
Here are some suggestions:

Each time SETI@home launches, it optimizes itself for the current monitor color depth (256 colors, Thousands, Millions). If you change the color depth while SETI@home is running, it may slow things to a crawl.
It should run reasonably at any screen depth, but it will run somewhat faster at lower screen depths (256 colors) than higher ones. Screen resolution (800X600, 1024X732, etc.) should have negligible effect on speed.
In screensaver mode, SETI@home runs fastest if the screen is blanked, so for maximum speed set the preferences to blank the screen after a short time.
In application mode, SETI@home runs fastest if its window is collapsed, using the Windowshade feature under OS 8.0 and later.
Many Macintosh models are equipped with the Energy Saver Control Panel. This can put the computer's processor, hard drive, and monitor in a low-power sleep mode if the mouse and keyboard have been idle for a period of time. Although SETI@home can continue to run when the computer is in sleep mode, it will run much slower. You may want to change the settings in the Energy Saver Control Panel to let SETI@home have more time. Some people prefer to have the computer never sleep so it can analyze more signals, while others prefer to save electricity. The choice is up to you.

Will SETI@home run faster with more RAM (e.g., 256 MB instead of 128 MB)?

SETI@home uses about 16 MB of RAM while it's running. Beyond a certain point (typically 64MB, more if you run memory-itensive applications) more RAM won't make it run faster. so you see you dont need or even use your 128 megs of ram ( or in your case 120megs)


Enjoy /forum/biggrin.gif

Mntsnow

ps. I was not trying to belittle your system (my daughters machine is based on a pc-chips mobo) (it used to be MINE utill I could afford to build my "performance" system) Just making a informed and well researched comment. /forum/smile.gif

[This message has been edited by Mntsnow (edited 10-20-1999).]

[Axel]

AMD chips, except their newest one, have an issue with floating point calculation - I.E. - they're slow. Seti will only use about 16 MB ram, I've got it running on 3 different boxes, so I doubt it's the RAM.
Have you checked you motherboard jumper settings to make sure the CPU voltages are correct? Have you benchmarked the processor? Hey - perhaps it's a bad processor - it happens.

[deep_sky]

Axel...
yes I have benchmarked the processor 300 ways to sunday and it appears to be running normally. However, I just benched the motherboard and ...ERRORS.
CMOS ram, clock ticking, and alarm are failing the test. Could this make my comp a little slower than it should be?