Cold boot failures x58A-ud3r Rev2 bios FB

[hello people]

Funky...have a read of my last few posts.........I'm a long way from breaking out the Bollinger. And I clearly stated that.

XMP has been working flawlessley for me. Since I installed the new ram and set bios I have had 2 successful cold boots, each cold boot about 3.5 hours wait. I will be nervous in the morning for a cold boot though.

We'll see

[Dark Mantis]

FunkyErn

One other thing to maybe consider is the Noctua NH-9UB cooler - some other forum postings suggested reverting to the stock Intel cooler fixed similar problems. This did not make sense to me so I have not yet tried that. What does everyone think? Worth a try or not?

I can't see any correlation between the heat sink and the problem except for if it was caused by the CPU not seating properly or something and then moving it to replace the heatsink could make a difference, but I think it is unlikely. As I said earlier I think it is more likely a problem that is BIOS related.

Dark Mantis - the TR3X6G1600C9 is on the Corsair list for the GA-X58A-UD3R - I did check that before ordering the parts. http://www.corsair.com/configurator/product_results.aspx?id=1362102

Yes I know the Corsair memory is on the QVL. Nobody was disputing that fact.

[Dark Mantis]

Funky...have a read of my last few posts.........I'm a long way from breaking out the Bollinger. And I clearly stated that.

XMP has been working flawlessley for me. Since I installed the new ram and set bios I have had 2 successful cold boots, each cold boot about 3.5 hours wait. I will be nervous in the morning for a cold boot though.

Try putting the system in the fridge with the Bollinger and you can kill two birds with one stone.

[hello people]

If only I had some Bollinger....and a fridge...

[Gary123]

I am trying to run it at 1066 and not the full 1600. This does not make a difference.

Ok, but if the board is not advertised to support DDR3-1600 modules in general, is it necessarily expected to support DDR3-1600 modules running at 1066? (I don't know whether DDR3-1600 memory is required, or advertised, to be backward compatible with DDR3-1066, although it does often seem to work that way.)

It's tricky to say what (if not incompatible memory) distinguishes your system from others that work properly. It could, as you suggest, be a problem with the FB BIOS, but not necessarily, since many of us use FB successfully. Possibly there is an interaction between different (but correct) BIOS versions and some other problem.

One other thing to maybe consider is the Noctua NH-9UB cooler - some other forum postings suggested reverting to the stock Intel cooler fixed similar problems. This did not make sense to me so I have not yet tried that.

Have you looked at your core temps while the system is under load?

[Dark Mantis]

What you have to take into account is manufacturing tolerances on every component that is used to make the motherboard. These tolerances can add up or cancel each other out and that is why no two boards behave exactly the same, the same as CPUs, memory or any other device.

[FunkyErn]

Mmmm... this is all very strange and irritating....

I flashed FC1 bios with XMP enabled at 1600 and ran OK for 30 mins and then the OS just froze. So I have reverted back to the FB bios, loaded fail safe seetings and then enabled XMP profile1 at 1600. This has been booting fine (warm) and running without any BSODs - am just running a SiSoft Sandra burn in test on it. However, the true test is the cold boot from an overnight "rest".

Hello People - what model number is the Corsair 1600 XMP memory you are currently using? The TR3X3G1600C8D? Let us know how it goes in the morning.

Dark Mantis - I am inclined to agree with you on the question of the heatsink. It does not make sense. After reseating the CPU and replacing the cooler I took care that it was not too tightly attached incase that was flexing the board or socket in anyway. It makes sense that no two motherboards will be exactly identical.

Gary123 - As far as I am aware from what I have read about DDR3 the faster speed ones are backwardly compatible. The memory timings are different for each speed setting though. Regarding CPU temp - the average core temp is around 37-39 degrees centigrade. I believe this is acceptable.

I guess the acid test will be in the morning once the pcs have completely cooled down overnight. Fingers crossed...

[Gary123]

Regarding CPU temp - the average core temp is around 37-39 degrees centigrade. I believe this is acceptable.

Yup, better than my temps anyway. With the memory at 1600, CoreTemp shows 37-40° at idle, 54-62° at 50% CPU load, and 68-71° when the temperatures level off during the Prime95 torture test. (i7-950, no OC, CoolerMaster Hyper 212+)

[Dark Mantis]

Yup, better than my temps anyway. With the memory at 1600, CoreTemp shows 37-40° at idle, 54-62° at 50% CPU load, and 68-71° when the temperatures level off during the Prime95 torture test. (i7-950, no OC, CoolerMaster Hyper 212+)

Maybe you could do with a little more cooling in your case or even just a better ballanced airflow.

[Gary123]

Yup, better than my temps anyway. With the memory at 1600, CoreTemp shows 37-40° at idle, 54-62° at 50% CPU load, and 68-71° when the temperatures level off during the Prime95 torture test. (i7-950, no OC, CoolerMaster Hyper 212+)

Maybe you could do with a little more cooling in your case or even just a better ballanced airflow.

Perhaps, though it's a Gigabyte 3D Aurora--a large case with three 120mm fans (in addition to the CPU fan).

Am I correct to understand that my core temps, though higher than average, are still safely within specs? (During the Prime95 torture test, with 100% CPU use, the processor was still running at the turbo speed of 3.2 GHz, so the processor itself doesn't seem to think it's too hot.) Memtest and Prime95 haven't shown any errors, and the system has been completely stable so far.

[Dark Mantis]

Yes you seemed to stay within the saftey zone ok. It's just that you have to remember that if your system is nice and clean and everything working at optimum power it can only get worse. As dust etc builds up so does the temperatures.

[Gary123]

Yes you seemed to stay within the saftey zone ok. It's just that you have to remember that if your system is nice and clean and everything working at optimum power it can only get worse. As dust etc builds up so does the temperatures.

Yup, I'll keep an eye on that. But for now, I'm happy to declare victory and stop spending time on this. Plus I don't want to keep tinkering until I eventually break something.

[hello people]

Funky I'm using Corsair CMX6GX3M3C1600C7 2gx3

By the way........just had my first successful morning cold boot in 1.5 weeks. That's 3 cold boot wins in total.

Still not counting chickens though.

Good luck everyone

[Dark Mantis]

Well it sounds hopeful now and you have to crack the Bollinger at some point!

[FunkyErn]

Hello People - Congratulations!! That must be a relief 

Dark Mantis - my cold boot issues are still unresolved.

I ran Sisoft's burn in for 3-4 hours on Sunday without any issues on bios FB. Monday morning BSOD BCCode 1A on first boot after overnight cool down! I then decided to swap the Noctua cooler for the stock intel retail package one - I wanted to eliminate all options inspite of us both thinking this would not be the issue. Booted first time without any issues (of course it would now that is was warm), I then ran SiSoft's burn in for about 3 hours and no issues. However on first cold boot this morning the BSOD BCCode 1A stuck again as soon as I logged in to Windows. So I think we can now safely conclude that the cooler is not the issue. The problem manifests itself only for about 15 mins on cold booting from overnight. I do switch off my computer at the mains overnight - could this impact things?

I have now flashed the FC1 bios and will sit back and see what happens overnight. When I tried this previously Windows stopped responding after 30 mins. I have switched from using my wireless mouse and keyboard to my old wired ones to eliminate any other variables. Any idea where I can get the latest bios (FC2?) from?

Gary123 - the temps I quoted were at idle. While running the burn-in tests the core  temps varied from about 40 at idle to 73 at peak load.