New BETA BIOS 0902 out

[Juanlo_]

Thanks to Raja and the others for their comments, I'll return this memory kit, purchase a G.Skill Flare X when they are available and wait for the enhanced memory support that AMD will release in May so motherboard makers will incorporate them into the BIOS. I'll be running at 2133 Mhz for now.

[StarfireX]

So i turn my system on today and now it will not boot with any custom overclock settings at all. Even though just last night i had it running perfectly fine with a 3.8 overclock 110 BCLK, increased voltages for CPU and DRAM. Everything was fine. I even cold booted it a couple of times and it booted every time.
But i switch it on this morning when i wake up and BAM, nothing. Stuck on error code 55. Other times it will stop on random codes like 0d and 8, but most of the time it's 55....
I seriously have no idea what is going on. Is this motherboard sensitive to daylight or something?

I have this same thing happen to me, one day boots fine over and over, next day I have to reset the bios. *If I enter my setting again it will be fine again, no idea, this is on 0902

[Kobura]

I have this same thing happen to me, one day boots fine over and over, next day I have to reset the bios. *If I enter my setting again it will be fine again, no idea, this is on 0902

Highly likely this is depending to some temperature. I had this as well, but got it stable by putting DRAM Boot voltage to my DRAM voltage and VTDDR to around 0.8V.

[AndehX]

Highly likely this is depending to some temperature. I had this as well, but got it stable by putting DRAM Boot voltage to my DRAM voltage and VTDDR to around 0.8V.

I have these same voltages set, and yet this problem persists. The temperature thing is interesting though. It's the only thing I can think of that is different from last night. My system has been off all night, so its physically cooled right down. It seems to overclock better when its physically warmed up.

[Raja@ASUS]

Memory training from cold BOOT and warm cycle is different. This is due to the effects of temperature on signaling. Passing one way doesn't mean you'll be guaranteed to pass the other. If you can't tune it out, downclock the memory and wait for the next update. Will save you tearing your hair out and needing to keep a daily diary of what the system does.

[MNMadman]

You could also try increasing the SOC voltage. Mine reboots fine and also cold boots fine (even when off overnight) but my SOC is at 1.155v. At lower SOC it wouldn't always cold boot the first time.

[AndehX]

Memory training from cold BOOT and warm cycle is different. This is due to the effects of temperature on signaling. Passing one way doesn't mean you'll be guaranteed to pass the other. If you can't tune it out, downclock the memory and wait for the next update. Will save you tearing your hair out and needing to keep a daily diary of what the system does.

I have never heard of this, nor encountered this on any other system I have ever built.

You could also try increasing the SOC voltage. Mine reboots fine and also cold boots fine (even when off overnight) but my SOC is at 1.155v. At lower SOC it wouldn't always cold boot the first time.

tried it. Doesn't work.

[AndehX]

Unbelievable! It seems that the physical tempreature of the system really does have an effect on the OC. I've just gone into my bios now (system has been on for about 4 hours now) loaded my 3.8ghz profile, and just like that, it boots first time! Utterly mind blowing. I've never encountered this on any previous system I've built. I really hope this is something that can be fixed in a bios update, because it's incredibly inconvenient.

[Praz]

I have never heard of this, nor encountered this on any other system I have ever built.

Hello

This is a property that is true of all electronic circuits. A properly designed system has margins that take into account thermal drift and noise. The further the system is run outside of its specifications the smaller this built-in margin becomes.

[AndehX]

Hello

This is a property that is true of all electronic circuits. A properly designed system has margins that take into account thermal drift and noise. The further the system is run outside of its specifications the smaller this built-in margin becomes.

While that may be true, the fact still remains that I've never had to deal with this issue until now. And I've been building and overclocking systems for over 15 years