Well, after a couple of problems I ran memtest and it comes up with a lot of errors. Orthos works fine on blend, small and large FFT. I've tried reducing my overclock, relaxing the RAM times and both decreasing and increasing the RAM voltage slightly. All just seem to result in memtest finding more errors more quickly! Any ideas on other stuff to try, or am I just gonna need to replace the RAM?
memtest errors
- Thread starter sidefxv1
- Start date
TBH, I'm not sure I ever did run memtest on it before. 
However the problems that led me to use memtest (CRC errors when extracting rar files + other stuff) have only started a few days ago. Before that everything was peachy. RAM is the system in my sig.
RMAing might be a problem since I got it in the UK but am living in New Zealand at the moment. doh.
However the problems that led me to use memtest (CRC errors when extracting rar files + other stuff) have only started a few days ago. Before that everything was peachy. RAM is the system in my sig.RMAing might be a problem since I got it in the UK but am living in New Zealand at the moment. doh.
hey netby, they've lasted pretty much since I joined OcUK, so since May. But most of the time they were 1000MHz @ 5-4-4-12. Which is why I'm not too fussed about RMAing them... they've lasted pretty bloody well as is. 
Odd how the P5B doesn't allow mem voltage to be set to 2.2V exactly though.
Odd how the P5B doesn't allow mem voltage to be set to 2.2V exactly though.Yeah, I'll have to find out what postage costs would be. TBH, I think it probably won't be worth it - at the very least I'd probably end up waiting weeks.
On the plus side I tried both sticks one at a time and it looks like only one has died - the other gets a completely clean bill of health from memtest. So, I'm just running with the one for now.
Interesting how orthos seemed to pass pretty much everything, but memtest failed within literally 1 or 2 seconds of starting to run. Think that'll have to be my new first step in stability testing a clock. Probably should have been already... but I always thought orthos\prime95 would pick up that sort of thing...?
On the plus side I tried both sticks one at a time and it looks like only one has died - the other gets a completely clean bill of health from memtest. So, I'm just running with the one for now.
Interesting how orthos seemed to pass pretty much everything, but memtest failed within literally 1 or 2 seconds of starting to run. Think that'll have to be my new first step in stability testing a clock. Probably should have been already... but I always thought orthos\prime95 would pick up that sort of thing...?