Re: mp3 encoding - lame?
jurriaan wrote:
> I did run memtest86 after overclocking, no problems during a 24 hrs
> period. The I compiled 250 kernels, which were all identical. Still,
> lame is obviously very sensitive.
> The moral of the story: overclocking a 3700 San Diego to 2750 MHz works
> most of the time, but not with lame. For now, 2640 works a lot better.
>
>
>>- Check to make sure your system is being cooled adequately.
>
>
> the cpu is never above 45 degrees C.
>
>>- Run memtest86 and see if it likes your RAM.
>
>
> 24 hrs - no problems.
>
> This must sound lame, but I really thought I had it stable, with memtest
> and compiling kernels.
No, it's only LAME once you get it working right. :)
Kernel compilation might not be so good a test since it's heavy in I/O
-- your CPU isn't working as much while it's waiting for the hard disk
and cools down somewhat. I've mostly used mencoder for stability testing
in the past. Burnk7 (from the i386 cpuburn package) is very good for
heating up the CPU, though, and I might start using that instead of
mencoder. Bzip2 seems to work well for testing overclocked RAM, but I
try to stay away from that.
-Corey
Reply to: