PDA

View Full Version : "internal computation error"



Mystwalker
11-04-2002, 04:50 PM
I just saw through the log file and spotted this little fella two times:

internal computation error [excessive roundoff]! check your memory/processor. test will restart in 5 minutes.

Doesn't sound got to me... :cry: :help:

Ah, and another thing: When I set the max log file size to 0, that means it can take as much space as it wants, right?

jjjjL
11-04-2002, 08:34 PM
you are the first person i even heard of having this error since i added a message to alert the user of it. there are other messages with less dire ramifications, but the excessive roundoff is almost surly a hardware error. i've induced it on purpose a few times, but i had to do terrible things to my system... like remove my cpu fan from the heatsink for a few minutes. yourha system may need better cooling, even if it's not overclocked. that error really wouldn't occur (esp several times) on a stable computer.

-L

Mystwalker
11-05-2002, 04:32 AM
Well, that system is overclocked, but not as much as it's used to be. Hm, gonna fiddle around a bit though.
Is it sure that all computing errors get intercepted? If not, I would stop the program on that PC. Don't want to be the guy whose computer misses a prime number and falsifies the sierpinski problem... ;)

prokaryote
11-05-2002, 05:40 AM
Had the same error when I pushed the OC too much. Occurred as the computer farm room got too hot. The client would pause for 5 minutes and do okay for 1 or 2 blocks then the error would show up again. I backed off on the overclock and all was fine. As the room cooled (night time) I could push the OC again.

smh
11-05-2002, 10:35 AM
Originally posted by Mystwalker
Well, that system is overclocked, but not as much as it's used to be. Hm, gonna fiddle around a bit though.
Is it sure that all computing errors get intercepted? If not, I would stop the program on that PC. Don't want to be the guy whose computer misses a prime number and falsifies the sierpinski problem... ;)

Try the Torture test of the GIMPS (http://www.mersenne.org) client. If that one gives you the errors too then you definately have a hardware problem (running at the rated speed will probably resolve that erroro though)

smh
11-05-2002, 10:42 AM
Originally posted by Mystwalker
Well, that system is overclocked, but not as much as it's used to be. Hm, gonna fiddle around a bit though.
Is it sure that all computing errors get intercepted? If not, I would stop the program on that PC. Don't want to be the guy whose computer misses a prime number and falsifies the sierpinski problem... ;)

Try the Torture test of the GIMPS (http://www.mersenne.org) client. If that one gives you the errors too then you definately have a hardware problem (running at the rated speed will probably resolve that erroro though)

Mystwalker
11-05-2002, 04:55 PM
Hm, in a certain way, the temperature is too low. :eek:
It stays at a constant 41°C/105°F (+x coz the Duron has no internal temp sensor :( ). I think it's the CPU voltage. It's set it to 1,625V, which wasn't a problem so far - but I'll up it a tad soon and watch if it will make the problem go away...