Jump to content
Corsair Community

Bit errors with TWIN2X4096-6400C4DHX, not showing up with memtest86+


bcc

Recommended Posts

I have 2 2GB sticks of TWIN2X4096-6400C4DHX memory in my system, a gigabyte ga-p35-ds3p, and I'm experiencing occasional undetected data corruption. The corruption occurs when running the motherboard at stock speeds and the memory at stock jedec timings and voltages. Problem first exhibited itself as data corruption on disk writes, approximately 1 bit error every 100 GB of data. I was able to simplify the problem to just copying and verifying all-one-bits across memory, no disk I/O involved. In this case I see approx. 1 bit stuck at 0 for every 100GB of copying. Problem happens with just 1 of the 2 dimm sticks only (or when both are used).

 

The strange part is memtest86+ does not report errors when I leave it running overnight, even when I just run test#5.

 

Should I just RMA these?

Link to comment
Share on other sites

Ok, it took me a couple days to get the testing done but I did manage to reproduce the same bit error with the 1 problematic DIMM on another system (different motherboard, cpu, power supply, enclosure, et all). Oddly I haven't been able to reproduce the problem with windows, and under linux, the problem only turns up some of the time.

 

That is, some times I can run a test for 8 hours and not get a failure, some times it takes just a few minutes. It seems I have to run other background tasks on the system to warm up the case a bit (get the motherboard to perhaps 40C) before the memory errors start to occur.

 

I now suspect I haven't seen the problem with memtest86+ because the motherboard stays cool (<= 35C) when it is the only thing the system is doing.

 

Time to RMA yet? I would hate for corsair to just test the stick at cool temperatures, pass it, and just send the pair back to me.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...