X9000 on my M570RU... just arrived-test

Hi,
today is arrived my X9000 ES with final stepping C0/M0 (Q174).
The first thing I've done is stressing @ 2.8Ghz and default Voltage > OK @ 65C
Second @ 3.0 Ghz and default voltage > OK @ 77C
tird @ 3.2 Ghz and default voltage > Errors in Orthos or Everest after few seconds (without shutdown or blue screen) but games and benchmarks work without problems!!


Now I will continue on testing but why it doesn't work properly @ 3.2??
It should work @ default voltage or not???
Any experiences??
What about the voltage limit to 1.200 V in bios??

My main bios is 1.00.08 and KBC 1.00.09S

 

Anyone with the same experience?

 

Default Voltage are 1.313/1.325/1.338V @2.8/3.0/3.2GHz.
I also get Errors in Orthos @3.2G......

 

Oh finally, the man with a Montevina "on" a Santarosa!!
I use RM CLock for the voltages and it seems to be 1.2250 @ 3.2Ghz..
Do you use CPU-Z instead??

So it is normal that errors appear in orthos or everest, right?
I am very relieved!!

Another problem:
when I change the management profile in RMClock and then come back to "no management" the CPU is not more able to reduce Frequency @ 800 Mhz,
instead it remains on 1.20 Ghz.
Could you make a photograph of your RMCLock configuration??
(I have FID and VIDS unlocked)

Which bios version are you using?

sorry for questions

 

I get no errors at 3.2GHz...

PCWizard 2008 reports 1.338V

Newest bios for m570ru is 1.00.11

I have B1 stepping

 

Really no errors?? It's strange..

What about forcing it to run @ 800 Mhz and 0.9250??

 

It'll run at 800MHz if I allow windows to "power save" to 25%, but the standard voltage for 800MHz is 1.113V, I haven't bothered to undervolt

Interestingly, though PC Wizard shows it running at 800MHx, CPUZ still shows 1.6GHz...

 

CPU-Z doesn't recognise SLFM, whereas PC Wizard and Vista do....

Also try using CPU-Z v1.41, since the later versions of CPU-Z are buggy with voltages....

 

Ok, but how to force it using RMCLock ??

With T9300 8x corresponded to SLFM and so selecting it I had no problems.
It reduced the FSB to 100Mhz > 800 Mhz

With X9000 8x corresponds to 200Mhz FSB> 1.600 Mhz
X9000 & 6x corresponds to 200Mhz FSB> 1.200Mhz
This is the minimum I achieve forcing management..but I would like to see it with FSB@ 100Mhz x 8 @0.925v for the battery power!

Thanks for helping

 

up
Anyway to force it @ 800Mhz??
Theriko can you explain better?
Thx

 

Edit: You can force 800MHz using the windows power management by setting maximum processor state to 25% as shown here running orthos at 100% cpu load at 800MHz

 

Any idea to do this in XP??
However thanks

 

See edited above post

 

in XP??
there isn't so specific option and enabling "power saving" from xp produces nothing>1.20Ghz

 

Not a clue about doing it in xp I'm afraid

 

However 3DMark06 with system in sig 11696 !!

In crysis warhead @ 1440x 990 = 0.40 FPS more so almost NOTHING!!

In Far Cry 2 also almost nothing more than T9300 !!

 

Here it is my latest 3DMark score 11753
and Vantage 5993

 

Hmmm... seems that proccy is not of much use in games!!

Though u'd notice a considerable improvement in proccy intensive tasks such as video encoding nd stuff!!

Mine with the T9300 Oced to 2.8Ghz + stock 8800m GTX easily gives 35fps in FarCry 2 at 1600x1200, all - very High, no aa!!

 

38.5 fps in FarCry2 1600x1050 Very High DX10 Vista 64..
I'm gonna try it on Windows 7 later

 

my b stepping has no errors my c stepping errors out.. strange...

 

And this is very strange!! The final stepping shouldn't have these problems..
and these ARE problems because Orthos says that in its operations the cpu returned a value different from the real result!
However in game and encoding there are no problems at all, and also with orthos but @ 3.0Ghz it gives no errors...
It could be interesting to see what a NON-ES do with ORTHOS!!

 

mine had no errors olie, but i could only over clock from 2.8 to 3.0.

try dropping the voltage when you over clock at 3.2

 

I am @ 3.2Ghz and 1.1875v (from RMCLock)..lowering voltage BSOD,
raising voltage still crashes in Orthos but in other ways is stable!

 

i mean lowering the voltage a very tiny bit. not to it's full drop voltage. because it wont have enough to run the chip. so if the normal was 1.250 then you would try 1.200 or something close to that.

 

Anything from 1.35>1.250> to 1.18v cause Orthos to crash...but everything either works fine!
1.1875 is stable, 1.17 not very stable

 

on the next crash, can you take a picture...maybe i can get an idea if i see exactly what you see... also run a temp monitor next to it as well.

 

temps cannot increase because crashes after few seconds..

 

run that test as small fft's and not blend.
i know this isn't the same, but it is an es x9000 all the same.

 

here it is

 

try prim95 if you have it...

 

wprime works without problems..