Memory testing software??

  • Thread starter Thread starter half_pint
  • Start date Start date
H

half_pint

Anyone know of a free memory testing software.
I want one which I can basically leave on and
'hammer' the memory for a long time to see if
it produces any results.
The only ones I could find just seem to do a quick
check.
Also preferably not one ran from a floppy (DOS) if possible.

Thanks half_pint.
 
half_pint said:
Anyone know of a free memory testing software.
I want one which I can basically leave on and
'hammer' the memory for a long time to see if
it produces any results.
The only ones I could find just seem to do a quick
check.
Also preferably not one ran from a floppy (DOS) if possible.

Thanks half_pint.
this one will run for over 24 hours i believe:

http://www.memtest86.com/
 
Memtest86... will need a boot disk, but it's the best out there.

Check the options to enable some of the longer, more stringent testing.
 
Yes thanks to all. I have it and I have ran it now with no problems so far.

There's also a newer fork, called MemTest86+ at
http://www.memtest.org/ which works with the newer
motherboards.

Other programs which are good at finding marginal
hardware are QuickPar and Prime95. Prime95 has been a
notorious way to stress-test your CPU/memory for quite a
few years. There's even a menu option that lets you run
it as a stress test then as a Mersenne.org client.
 
Alien said:
.... snip ...

I've had several systems where memtest86 passed the RAM as
faultless yet Win2k or XP crashed with what turned out to be
memory errors. So I don't think its perfect.

However ECC memory is perfect, to all practical purposes. If you
don't insist on having it in your machines, the result is entirely
your fault.
 
Alien Zord said:
I've had several systems where memtest86 passed the RAM as faultless yet
Win2k or XP crashed with what turned out to be memory errors. So I don't
think its perfect.


I ran it OK for 10 hours, maybe those systems were under powered too?

How do you know it was memory? With inadaquate power would not the
memory *appear* faulty anyway???
 
However ECC memory is perfect, to all practical purposes. If you
don't insist on having it in your machines, the result is entirely
your fault.

I agree. But, I can't for the life of me get the Intel Desktop D865GBF
mainboard to run with ECC memory.
 
I ran it OK for 10 hours, maybe those systems were under powered too?

How do you know it was memory? With inadaquate power would not the
memory *appear* faulty anyway???

For me, it was RAM timings (CL 2.0 vs CL 2.5).
MemTest86 would check out 100% (running for 10 hours at
a shot), but the system would still be unstable in
normal use when 2 DDR DIMMs were installed at the same
time. System was perfectly stable with 1 DIMM, but not
2. Prime95 and QuickPar both were symptomatic of memory
problems (as well as lots of random, inexplicable
reboots).

Underpowered systems typically manifest themselves in
other ways (hard drives that turn off, RAID arrays that
lose drives).
 
Back
Top