Help decoding C1 error

[EtherealN]

Twice now I've had BSOD's caused by an error code 1a, whereafter boot fails on POSTcode C1 (base memory detect). However, since these two instances are a spaced by a lot of time (2 weeks I think) it has now been very hard to replicate and I haven't found any information on the net that helps decode this.

 

In each of these cases turning power off completely and then starting the computer normally works flawlessly.

 

So if anyone has any ideas I'd be most grateful. :)

[Panzertard]

There's a message on the XFX site;

"We recommend updating the BIOS on your XFX Nforce 780i SLI ( MB-N780-ISH9 ) to version P08."

Other than that, one have to register to get access to the support (manuals) for the XFX mainboards.

You wouldnt happen to have it in PDF? To be able to check the boot fail codes in details?

[EtherealN]

Done that (overclocking is impossible otherwise on that board, since it faults when unlinking memory), and the sum total of information the manual gives about that fault is that that stage is Base Memory Loading.

[Slayer]

You are probably overclocking your memory too much PC2-6400 is 800 mhz...thats like a 36% overclock. at 1092 mhz your overclock is higher than a PC2-8500 speed(1066)mhz. I'd back it down to 1066 if you can your timings look the same as for the PC2-8500. I used to have some Corsair Dominator that wouldnt overclock even 100mhz higher.

 

Overclocking memory does nothing for you anyways because you have to loosen up the timings so even though its clocked faster it has higher latency. Even with the most expensive memory theres litterally zero difference in FPS in games compared to lower speed memory with faster timings.

[EtherealN]

Second time it happened there was no overclock on the memory. After that one I re-linked the memory sticks (giving a more modest 1020MHz) and so far it hasn't happened again, but since it was infrequent that's no guarantee. But it's not the overclock itself that did it (then the second instance shouldn't have happened unless there's physical damage to the sticks, which is a possibility), but there might be something funny in the FSB handling unlinked memory, especially since the previous BIOS version on the MB would freeze up and die if you unlink.

 

As for the gains, there actually are gains to it, though the only tests that reliably show difference are synthetic ones - 4% performance on CPU tests in 3DM. (DCS specifically still has the CPU as bottleneck.) I still have my almost year-old stats from the first test-runs up at this adress: http://bluefire.dnsalias.com/~tats/files/art.htm

 

But as for too much overclock - it passes Prime95 memory torture tests with flying colors, so it doesn't give faults at least. But hey, probing the limits is what it's all about. :D

Edited December 26, 2009 by EtherealN