Occt avx2 ошибки

If it does it at stock settings it could be a few reasons:

1.) RAM needs more voltage, check the specs to be sure.
2.) RAM is bad.
3.) CPU is bad/needs more voltage.
4.) Chipset is failing/needs more voltage.

Get it passing the test at stock before you try overclocking.

First off have you checked your voltages ? RAM , Chipset , CPU …
The OCCT is a fine tester but I have had this issue with it and with prime95 before . Some say they are the only way to test for stability but if you are at stock settings and it crashes then that should tell you some thing about the program it’s self . I don’t trust them all that much at all . your better off trying things like Intel burn in test or running 3dmark06 2-3 times or any thing els . OCCT for me is more of a torcher test some thing that the CPU will never see in normal operation so is prime95 .

OCCT is a torcher test, but Intel Burn In isn’t? Intel Burn In is one of the worst conditions to put a processor under, nohting like most real world situations. Also, OCCT uses the same backend the Intel Burn In Test, you just have to select the LinPack test.

OCCT is a fantastic stress proggy, CPU, GPU, PSU

I use it for all of em, Linpack is the Intel Burn Test, but thats unnessecary IMO, just tests your coolers capabilities more than anything.

I say, 1 hr of OCCT:CPU on auto and ur good.

Though running at stock and still failing, either points to temps, RAM or NB volts being an issue if you ask me.

kenkickr doesn’t say what volts, so I assume he kept within recommended specs. It wouldn’t be the first time a stress tester has killed hardware. Theres many stories of Furmark killing graphics cards that were at stock clocks.

Furmark over heated the cards. It wasn’t the stress tester that killed the cards, but the poor build quality of the cards themselves, and the fact that the manufacturer let the cards leave the factory with cooling insufficient to handle heat generated by the cards.

I’ll work on it for ya

W00t!

If it does it at stock settings it could be a few reasons:

1.) RAM needs more voltage, check the specs to be sure.
2.) RAM is bad.
3.) CPU is bad/needs more voltage.
4.) Chipset is failing/needs more voltage.

Get it passing the test at stock before you try overclocking.

First off have you checked your voltages ? RAM , Chipset , CPU …
The OCCT is a fine tester but I have had this issue with it and with prime95 before . Some say they are the only way to test for stability but if you are at stock settings and it crashes then that should tell you some thing about the program it’s self . I don’t trust them all that much at all . your better off trying things like Intel burn in test or running 3dmark06 2-3 times or any thing els . OCCT for me is more of a torcher test some thing that the CPU will never see in normal operation so is prime95 .

OCCT is a torcher test, but Intel Burn In isn’t? Intel Burn In is one of the worst conditions to put a processor under, nohting like most real world situations. Also, OCCT uses the same backend the Intel Burn In Test, you just have to select the LinPack test.

OCCT is a fantastic stress proggy, CPU, GPU, PSU

I use it for all of em, Linpack is the Intel Burn Test, but thats unnessecary IMO, just tests your coolers capabilities more than anything.

I say, 1 hr of OCCT:CPU on auto and ur good.

Though running at stock and still failing, either points to temps, RAM or NB volts being an issue if you ask me.

kenkickr doesn’t say what volts, so I assume he kept within recommended specs. It wouldn’t be the first time a stress tester has killed hardware. Theres many stories of Furmark killing graphics cards that were at stock clocks.

Furmark over heated the cards. It wasn’t the stress tester that killed the cards, but the poor build quality of the cards themselves, and the fact that the manufacturer let the cards leave the factory with cooling insufficient to handle heat generated by the cards.

I’ll work on it for ya

W00t!

normal? i let it run 2 hours, was fine for around 1 hour in.

temps are great less than 65 celsius on all cores.

should i just ignore this?

When is the last time it was updated?

Get yourself hardware monitor by CPUID (CPU-Z creator)

Get yourself Blender and the Ryzen Module here follow those directions.

https://www.techpowerup.com/forums/…zen-blender-benchmarks-at-200-samples.228686/

Get 7zip and a large compressed file, set 7zip to utilize all cores and threads.

Get Unigen Heaven/Valley.

Get the Latest 3DMark tool.

1. Have Hardware Monitor on Screen.
2. Run Ryzen Blender with those settings. If it completes successfully post your Voltages and temperatures here. If it freezes then a voltage might be too low, for my 8350 it was vcore at one point.
Post hardware monitor screenshot.

3. Run 7zip and decompress a large zip or 7z file, then compress it, with max cores/threads. Have hardware monitor open. Post images. If it freezes, check settings.

4 Run unigen tests and 3dmark. No freezes you are ok.

If everything doesnt freeze/bsod, ignore occt.

Did the mobo turn on MultiCore Enhancement by chance? My Asus Z97 always turned it on if you touched the XMP option.

Yes, it is on. 4700 and no downclocking by default.

So do I just ignore the error?

@Solaris17 ty I will run it now. looks like I can just ignore the error message though, since i guess maybe it is normal if turbo boost is on

You said full stock default settings in the thread title so that got me wondering. Anyways your overclocking, have you tried without MCE? I only asked because you could running be into instability that OCCT is picking up on.

thanks mate. will turn off xmp, and MCE, and let it OCCT run over night, if no issues then i will turn on MCE but leave XMP off, see what happens. if clear then I know its the ram, if not then i will keep going down the list, and try no MCE but XMP on

That’s degradation at work friend. It will slowly want more voltage over a year or two. Just the price you pay for more mghz

doubt it, lot of factors involved. my 2500k was 4,8ghz 24/7/365 for over 5 years. and still runs great at same everything in bios. its stable now, i just was doing something wrong and/or the MIS turbo enhancement in bios wasn’t stable. when i say stock, i meant MSI stock which included the turbo boost to 4.7 no downclocking. msi just released a bios update stabilizing voltages too, so yeah its all related.

Not all Silicon is the same, if your chip requires even a small voltage bump over its VID that is the very definition of degradation.

I didn’t know that, thank you.

Yup hence why some chips can handle hard ocs and others can’t. The better the cooling you have on the cpu, vrms and even on back of board where the cpu socket and vrms sit, even where the nb/sb are the better it will survive.

What version is occt up to?

Also after doing a clear cmos you should load optimized defaults.

when you say VID is too high, is this what you mean? even tho my temps are good and everything is stable my VID is too high on all 6 cores? this is stock MSI bios btw with turbo boost on lol

please advise if i need to turn this off. none of my manual oc’s are stable.

if your chip requires even a small voltage bump over its VID that is the very definition of degradation.

No, it is not. Requiring a voltage bump for an overclock is not degradation.

when you say VID is too high, is this what you mean? even tho my temps are good and everything is stable my VID is too high on all 6 cores?

I wouldn’t worry too much about the VID, it is always read high on my 8700K, pretty close to what yours is, even though the vcore is only 1.25v. The vcore is what you should be worried about, the VID hasn’t really been worth looking at for casual overclockers in years.

And as long as your overclock is stable and your temps are acceptable, that’s all that matters really.

If your cpu is unstable at stock speed and requires a small voltage bump what would you call that then?

If your processor was unstable at stock speeds and requires a small voltage bump, and it is a new CPU or has always been like this, then I call that defective, not degradation.

Of course, his CPU wasn’t at stock speeds, it was overclocked from the very first boot, so that point is moot.

Degradation is when a processor is completely stable at a certain clock speed and voltage, but over time becomes unstable at that clock speed and voltage.

ty all cheers

I figured out the issue, I was running Beta BIOS for the mobo which «tweaks cpu voltage», I went back to the last known stable BIOS, and 0 issues… 5ghz -2 avx, 1.40v, prime 95 small FFT for 3 hours now 0 issues. and before that 3 hours of OCCT 0 issues. lol. my bad mates.

normal? i let it run 2 hours, was fine for around 1 hour in.

temps are great less than 65 celsius on all cores.

should i just ignore this?

Can you please test your CPU with other software as well? Such as IntelBurnTest and Prime95.

Can you please test your CPU with other software as well? Such as IntelBurnTest and Prime95.

ah this topic can be closed… as I said it was a beta MOBO bios…

and yeah i run prime 95 12 hours and AIDA64 and OCCT all 3-6 hours each multiple times, and a 24 hour run. its all fine now. temps do get to about 80ish celsius area at 5ghz -2 avx offset and 1.40 volt with 0.015 + offset.

which is fine, but in gaming temps rarely go past 65 celsius.

That’s degradation at work friend. It will slowly want more voltage over a year or two. Just the price you pay for more mghz

Huh, Bullshit x100000

It can be a lot of things and its definitely not cpu degradation after a couple of months with a mild OC like this one.

ah this topic can be closed… as I said it was a beta MOBO bios…

and yeah i run prime 95 12 hours and AIDA64 and OCCT all 3-6 hours each multiple times, and a 24 hour run. its all fine now. temps do get to about 80ish celsius area at 5ghz -2 avx offset and 1.40 volt with 0.015 + offset.

which is fine, but in gaming temps rarely go past 65 celsius.

Glad you found your problem nice purchase btw the 8700k is a beast.