dnstech Posted May 24, 2005 Share Posted May 24, 2005 Greetings, I have a VS512MB333 ram module which I believe is bad. When using the mprime torture test it will run error free for hours with a max FFT of 16 but if I push it out to 512 (which should be bigger than cpu cache) I start getting errors within the hour. I've tried moving the module to an adjacent slot, and have tested the harddrive out using the maxtor test suite . . . so while it could be a motherboard issue I believe it to be ram at this point. I believe if it were a motherboard issue, it would've shown in the mprime FFT 16 test. Please advise on acquiring an rma #. Link to comment Share on other sites More sharing options...
Corsair Employees RAM GUY Posted May 24, 2005 Corsair Employees Share Posted May 24, 2005 First please use http://www.memtest.org to test the module and can you tell me the make and model of MB you have along with the bios settings you have set for both CPU and memory? Link to comment Share on other sites More sharing options...
dnstech Posted May 24, 2005 Author Share Posted May 24, 2005 First please use http://www.memtest.org to test the module and can you tell me the make and model of MB you have along with the bios settings you have set for both CPU and memory? I ran the memtest86 program included on the Fedora Core 2 cd without errors which is when I moved on to the mersine prime torture test. The machine is one of 8 (4 with identical components other 4 have slightly faster processors, and different CD drives :biggrin: but basically the same). The motherboard is Chaintech 7VIF4 Via KM400 and the bus speed is set to 166 as it is on all the machines. This machine is the only one producing errors. These machines are part of a 12 node compute cluster with a Front-end server. The cpus are run at standard settings (ie not overclocked). Link to comment Share on other sites More sharing options...
dnstech Posted May 24, 2005 Author Share Posted May 24, 2005 I ran the memtest86 program included on the Fedora Core 2 cd without errors . . . I've swapped the ram on two of the identical nodes and am rerunning the mprime torture test on both. If the problem moved with the ram we can move on to the rma process, right? If the problem didn't move then corsair is home free :D: and I will continue to test other components :sigh!: Link to comment Share on other sites More sharing options...
dnstech Posted May 24, 2005 Author Share Posted May 24, 2005 If the problem didn't move then corsair is home free :D: and I will continue to test other components :sigh!: You are absolved . . . the problem continues Link to comment Share on other sites More sharing options...
Corsair Employees RAM GUY Posted May 24, 2005 Corsair Employees Share Posted May 24, 2005 O.K.. Well from what you have posted if the systems running the 333 FSB CPU did not have the memory set to DDR333 that may be the problem, as they will by default try to run them at By SPD or DDR400 and if you try to run the memory at DDR400 with a 333FSB CPU it will cause a bottleneck in the chipset and may generate errors that are not from the memory. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.