AMD PSCHECK for Bulldozer UPDATE: NEWER VERSION & INSTRUCTION HOW TO USE IT

[FlanK3r]

CHew: u have some last K10 beta? Il testing FX under LN2 this week...Thx.

[FlanK3r]

So...for better safe I have to try enabled core 1/2 and 2/2 (under LN2 and PScheck)???

[chew*]

So how many people have lost there cpu using ps check and overclocking/validating on cores other than 1/2.......

I just got a PM from christian ney asking me why his chip died using ps check to oc 3/4..........

Lets get a count of dead cpu's for the non believers......

[FlanK3r]

Uff, I was lucky...We tried cores in first module and then after in second module

[Christian Ney]

I confirm what Chew* said, we just used ps check once to max the cpu after doing it from the bios, wanted to try core 3/4 and cpu died, error 56 at post

http://www.ocaholic.ch/forum/index.p...&threadID=3845

Maybe not software related but still

No issue before using ps check and core #1,#2

[Xtreme Addict]

I have tried second FX (6100, really crappy max ~ 7 GHz) on 3th core (first two do ~ 6700 MHz, third - 7 Ghz) and nothing happened. CPU is still alive (LN2 @ full pot & 2.0v vcore). So till now I tried 2x CPU, both are alive and I am CPU killer so I doubt it's simply PSCHECK bug, I think those CPUs just die. Just like GT or SB...

[chew*]

I have tried second FX (6100, really crappy max ~ 7 GHz) on 3th core (first two do ~ 6700 MHz, third - 7 Ghz) and nothing happened. CPU is still alive (LN2 @ full pot & 2.0v vcore). So till now I tried 2x CPU, both are alive and I am CPU killer so I doubt it's simply PSCHECK bug, I think those CPUs just die. Just like GT or SB...

That would make sense......if I had managed to kill a cpu using cores 1/2 which I have yet to achieve and I have hammered alot of them......It only happens when you leave 1/2 at default or lower and try to ramp another core set.

Usually you get a crash and then no post situation after attempting valid.....I've eliminated board from equasion.....we killed one on jiga and asus.....so it either software or cpu's and the fact that I can't kill one with core 1/2 after many attempts means it other core pairs or software....

Also it tends to happen to higher clocking chips far more than crappy clockers.....all chips that did die were 8 gig chips or damn close....

[Barr3l Rid3r]

Can someone please post a link for version 3.4?

[dinos22]

is it dieing because you are shoving 2v and the CPU is just ing out or the software?

I had an instant death with a second chip we tested in london running 2 cores only via bios in windows sitting in CPUz with 1.8v and changing multi from 6.5Ghz to next bump with AOD. So is AOD also killing CPUs or are they :banana::banana::banana::banana::banana: CPUs and cant handle it all?

I was only using Core 0 and 1

[Christian Ney]

Andre ran his chip at 2.2v, his chip is not dead

[dinos22]

Andre ran his chip at 2.2v, his chip is not dead

im thinking some chips can handle it and others cant

the chip that massman has now is the chip we used in London which we tortured all weekend from high CPU clocking to 3DMARK11 8 core benching around 7GHz with very high voltage.

[The Stilt]

Chips with lower leakage can probably withstand higher voltage before getting killed.
I´ve seen over 15% difference in power draw between chips from the same production batch.

I would imagine the deaths happen because of "spiking" (CUs off, cycle reset -> CUs on during init -> CUs off after init).
Or has this happened with low voltage too?

[Christian Ney]

As I said, on my side I have BIOS code error 56 aka ''Invalid CPU type or speed'', does this mean cpu's microcode is dead ?

[The Stilt]

does this mean cpu's microcode is dead ?

Microcode missing would be 59h.
No clue what causes 56h thou.

[FlanK3r]

If I understand Dinos, maybe the problem is between combination AOD and PScheck. I use Ai SUite and PScheck. Firts in AI Suite only FSB jump, than shuw down this program from taskmanager. After this I run PScheck and play with it (If I want higher FSB, I msut close PScheck, run AI Suite again etc...).

[Hondacity]

please can you guys re-upload the latest version?

thanks in advance.

[Christian Ney]

@Honda,

Hosted at hwbot

http://hwbot.org/forum/showthread.php?t=32513

[Hondacity]

thank you so much chris!

[cbjaust]

Has anybody noticed with the CVF-Z and 8350 that in PSCheck half the P-states are missing? I had BIOS 1302 and I only had Pb0 - 4200MHz, Pb1 - 4100MHz, P0 - 4000MHz and P4 - 1400MHz showing. P-states P1 to P3 were missing. I updated to BIOS 1901 and at first all the power states were available and showing up in PSCheck but after changing settings in the BIOS it's back to just Pb0, Pb1, P0 and P4. really annoying. It there a BIOS setting that affects this?

M/B CVF-Z Bios 1901
CPU FX-8350
RAM G.Skill TridentX 1866MHz CL8 16GB
VGA HD7970
PSU Seasonic S12 Energy+ 520W

Any ideas?