790 corruption issues?

[sonofander]

for what its worth, I started with the evga and moved to the striker... They both have their pros and cons. I think with the new bios, the striker is less of a bad buy

[GAR]

I would stick with the eVGA/XFX reference boards for now.

[Roger_D25]

Your probably right GAR although like sonofander mentions, once the BIOS gets sorted this should be a serious board (fingers crossed)! With that said however I'm one the few people who hasn't really had a single issue yet (except for the steep learning curve), lol! Then again I bought my board used from a guy I trust and he had decent results with it and no major issues?

[gmcg]

FYI guys ... something to think about ... I have no clue if this issue has anything to do with corruption though ...

My 790i Ultra SLI behaves like it is an analog thing, not a digital. Everything is kinda random, today it is stable, tomorrow it fails the first test run being cold or fails after 5 hours run - as it wishes.

What is important - I can confirm my subjective opinion with objective test results ... they look pretty strange for me and may be a cause of instability issues.



Please take a look. I had been running Everest Ultimate test as normally, everything was fine and the temps with 100% 4 cores load were around 56-57C. I have decided to stop the test. Normally my temps after it go down to 42-43C, but in this situation they went UP to 65C ... I can guess that it was some kind of Everest Lavalys problem or a parasite oscillation on the CPU. Regarding Everest - it is quite doubtful, such as I have problems with Prime95 failures as well. Then I started full 100% load test again and my temps went DOWN from 65C back to normal 57C under load. I have stopped the test again - temps went up again ... but after some time they went down as it should be - to normal 42C ...
The next test start gave me the normal behavior of the board - up to 57C under load and back to 42C idle without any TEMP UP issue ...

This oscillation - or whatever loads my CPU cores - happens randomly, my guess is that it is a reason of instability, but it is only a guess.

[Roger_D25]

Unfortunetly I've never run this stress test from Everest (although I just bought an Ultimate license finally) so I might give it a go! With that said I'm not exactly sure what the problem is but I don't think its instability to be honest. It sounds like some bit of software is loading up the CPU cycles (either Everest of something else)? From the graphs it doesn't look like your CPU is throttling and besides your CPU all the temps look pretty consistant? Of course this is just my opinion so take it with a grain of salt but I can't imagine system instability is causing your CPU load to randomally change like this?

[Guacamole]

EDIT: I see how you installed it. I was doing it the other way, horizontal bases and vertical retention mechanism. Dam if I had just noticed that lol!
But, changing from Ultra120 Xtreme to IFX-14, would it make a big difference? It would take me some time to change this and since the Ultra 120 has 2 extra pipes I wouldnt like to change and see no improvement.



Have you tried moving a file to another HDD? What about using sfc /scannow?

Sorry I was away for a day but yes I am trying a large file transfer right now and sfc/scannow is clear I posted my results in the 790i thread but I'll have to get back to you about the file transfer...

[gmcg]

With that said I'm not exactly sure what the problem is but I don't think its instability to be honest. It sounds like some bit of software is loading up the CPU cycles (either Everest of something else)? From the graphs it doesn't look like your CPU is throttling and besides your CPU all the temps look pretty consistant? Of course this is just my opinion so take it with a grain of salt but I can't imagine system instability is causing your CPU load to randomally change like this?

No, it doesn't. First all my software was off to run a test (and it is normal), second - look at the throttle graph ... it shows 100% CPU loading while the test was actually switched off!
What can it be, if the temps went UP? Something worse (more loading) than 100% stress test? What kind of program can do that?
All other temps are pretty consistent during the test, that is normal.

[Guacamole]

This oscillation - or whatever loads my CPU cores - happens randomly, my guess is that it is a reason of instability, but it is only a guess.

Yeah man I am looking at that graph and listening to what you are saying and that is really weird.. I don't understand how the instability would randomly overload your processor, I mean it seems like something has to tell the processor to run (operating system, drivers, programs, etc..) Maybe there is something worng with the bus drivers, Nvidia drivers have been known to freak out a little have you tried uninstalling everything and trying it again?? That is very odd another question is does Prime do the same thing?? I mean after you stop it does it overload the processor and such?? Have you considered running multiple stress tests to try and isolate the incident like OCCT, Orthos, Prime, Nvidia Stress test etc.??

[Guacamole]

Here guys I posted a link to some unfortunate results that I posted on the 790i thread but basically with P1 enabled and good memory performance at 1800FSB I get corruption on 702, Exact same settings with P1 disabled poor memory performance but no corrutpion...

http://www.xtremesystems.org/forums/...34#post2992834

[gmcg]

Yeah man I am looking at that graph and listening to what you are saying and that is really weird.. I don't understand how the instability would randomly overload your processor, I mean it seems like something has to tell the processor to run (operating system, drivers, programs, etc..) Maybe there is something worng with the bus drivers, Nvidia drivers have been known to freak out a little have you tried uninstalling everything and trying it again?? That is very odd another question is does Prime do the same thing?? I mean after you stop it does it overload the processor and such?? Have you considered running multiple stress tests to try and isolate the incident like OCCT, Orthos, Prime, Nvidia Stress test etc.??

OK, when the circuit is unstable, it can start to oscillate by itself. The analogy is like you are a singer and have a mic in your hand, you direct it to the speaker, the system starts to scream - it is an example of positive feedback and a parasitic oscillation.

Such things should not happen in the digital world, as there are no analog amplifiers in it ... but when the digital circuit is unstable, it can behave as analog circuit - see the picture below:

Those oscillations around VTT as shown on that picture can become to be non-fading ... at some frequencies with some conditions, as I guess.
I had to run plenty of tests, but it is not easy to define the problem, such as those tests randomly fail ... how can you catch the moment of beginning of oscillation, for example? You need some specific equipment to do it, such as high frequency oscilloscope ... and where to find one working at 2GHz FSB? I was lucky to have that weird test result, no one digital tech will explain it to you ... logically I mean.

[gmcg]

Here guys I posted a link to some unfortunate results that I posted on the 790i thread but basically with P1 enabled and good memory performance at 1800FSB I get corruption on 702, Exact same settings with P1 disabled poor memory performance but no corrutpion...

http://www.xtremesystems.org/forums/...34#post2992834

It is a known issue, that is why P1 and P2 were disabled in P03 BIOS.

[pinto]

It is a known issue, that is why P1 and P2 were disabled in P03 BIOS.

And performance are really bad with that bios...
Personnaly i kept the P03R2 bios and i'm really happy with it ( P1 and P2 enabled)

[gmcg]

And performance are really bad with that bios...
Personnaly i kept the P03R2 bios and i'm really happy with it ( P1 and P2 enabled)

Exactly ... though I personally like P04 a bit more than P03R2, the last one can cause that corruption easily ... but you can get better results AFAIK with P03R2 using it correctly. And to use it correctly someone needs either to be lucky or know a lot about GTLVREF or be extremely patient in experiments, etc.

[gcardinal]

ASUS has posted new BIOS: 0703.
It can be downloaded from ASUS FTP server:
/pub/ASUS/mb/socket775/Striker_II_Extreme/StrikerIIExtreme0703.zip

There is no info about what is in new BIOS.

[Roger_D25]

What a guy Gcardinal, thank you very much!

[Guacamole]

New bios is working awesome for the Striker P1 enabled rock solid performance accross the board. 703 No corruption as far as I can tell..

[Roger_D25]

Great news, I was going to flash to it this-evning but decided to test some new watercooling equiptment instead. With that said however I'm flashing over first thing in the morning! Thanks for sharing Guacomole!

[saaya]

New bios is working awesome for the Striker P1 enabled rock solid performance accross the board. 703 No corruption as far as I can tell..

what fsb and mem speed and timings are you running?

[Guacamole]

what fsb and mem speed and timings are you running?

I am runnning:

FSB/Mem linked and synced
FSB/Mem = 1770
Multiplier = 9.5 (4.204)
Vcore = 1.4275 (water)
Memory = 1.92
PLL = Auto
VTT = 1.26
SPP = 1.46
MCP = 1.55
GTL's = -36, -40, Auto, Auto
P1 enabled, P2 enabled



I enabled P2 yesterday and it had no effect on my system for the record. These settings have yielded 5 hours prime stable, sfc/scannow safe, 3DM06 looped for about 2 hours.. I have to say though at times things are still sketchy we need still a better bios, this one (703) is way better but we still are not there yet..

[slim142]

BIOS 0703 sounds kinda good, im still testing 0701

[bingo13]

NVIDIA (through the reference board partners) will be releasing the P05 BIOS tomorrow to address data corruption problems on the EVGA/XFX boards. ASUS has another BIOS after 0703, but that one should be fine now if our information is correct about it.

[slim142]

NVIDIA (through the reference board partners) will be releasing the P05 BIOS tomorrow to address data corruption problems on the EVGA/XFX boards. ASUS has another BIOS after 0703, but that one should be fine now if our information is correct about it.

And your source is...?
And how do we know 0703 "somewhat" fixes corruption when the changelog doesnt mention it?

[Roger_D25]

Great questions slim142! I don't mean to second guess your post bingo13 but with so many people having corruption issues this is big news and until reading your post I had not heard anything about new BIOS helping with corruption issues? In anycase thanks for putting it out there!

[Guacamole]

NVIDIA (through the reference board partners) will be releasing the P05 BIOS tomorrow to address data corruption problems on the EVGA/XFX boards. ASUS has another BIOS after 0703, but that one should be fine now if our information is correct about it.

Okay so what I have to say is (through extensive research) 703 does not "FIX" the corruption issue entirley it reduces the size of the window where corruption occurs.. I still get corruption under certain GTL, SPP combinations regardless.. It is still a work in progress so we are not there yet with 703 but big improvement over the previous bios's.. Anyway, I hope the ref. board bios helps their issues with corruption but I imagine there will still be underlying issues....

[fadetoblack]

NVIDIA (through the reference board partners) will be releasing the P05 BIOS tomorrow to address data corruption problems on the EVGA/XFX boards. ASUS has another BIOS after 0703, but that one should be fine now if our information is correct about it.

awesome! hopefully your source is correct..