>>> P4m 1.7 <<<

[jpetermann]

Looking good Thugs. Wish I could say the same for mine. See my thread for a list of thigns I am gonna try tonight.

PJ

 

[THUGSROOK]

i just completed M86 and this stick of ram aint scottish, so it must be CR@P!

(and it is) :disgust:

no fear ~ ive got 3 other sticks

 

[Leo V]

Amazing results so far...

...but I recall, i845E running at 200+FSB means you set memory to run at 1/2 the fsb bandwidth. However, with GBay you should have dual memory channels with 1:1 ratio. If we could only test 880fsb with dual 440mhz ddr channels, I have a feeling we would see much higher performance! (>7gb/sec theoretical bandwidth!)

(If Gbay could handle such fsb, holding fingers crossed!)

 

[jpetermann]

Leo V


Well, so for aI am unable to get past 175 fsb on my Granite Bay!!! Yes that is right, I did not even break 200!!! Still working on this thing though.

PJ

 

[Leo V]

What a bummer! This seemed to be where its potential would be really tested.. And you're using xms3500, that really shouldnt hold you back... I wonder if Asus p4g8x produces any different results...

 

[jpetermann]

Leo V,

I am not sure. If it is chipset, it won't matter what GB board you get. I am not ready to say that it is. I have a few other things to test, we will then have to see.

PJ

 

[Mikki]

I agree, way too early to call anything "cr@p" (except Thugsrook's first XMS3500 stick)....we still have a ways to go, different memory and boards to test, as well as eventual "work arounds"...

Hiya Leo, long time no see...

 

[Leo V]

Hmm... Kyle@hardocp got 180fsb out of his p4g8x... maybe it is a chipset limit?

I wonder if anyone tried the AOpen AX4R+ though--it seems to be "different" from other e7205 boards, and supports VCore all the way down to 1.1V. If low VCore works on the Aopen, at least it wouldn't overvolt the P4M chips, which would certainly reduce heat.

 

[THUGSROOK]

the XMS3500 has been removed from my system (garbage!) ~ now running a stick of XMS3000....

> Memtest86 < all 11 tests, 10 loops
passes upto 222fsb 444ddrb so far

(actually it was more like 35 loops since i let it run all night)

im gonna haveta redo all the damn benchies now :disgust:

 

[Mikki]

...at least you'll be busy for a little while longer....

 

[THUGSROOK]

 

[THUGSROOK]

sorry,
im getting sidetracked on this memory stuff.

back to work!

 

[ErikaeanLogic]

Originally posted by: THUGSROOK
another disapointing note:

while 2.7v and 2.9v vdimm can get you benchmarks ~ it aint gonna be 100% stable.

not one stick ive tried (XMS) can hold 2.7v+ and remain 100% stable.
itll even pass m86 ~ but give it something heavy and itll reboot :Q

ive said it before and ill say it again ~ 2.6v is the best youre gonna really do on a p4b533!

you think youre stable running 2.7v? try encoding DivX....

hmm, I haven't had that problem. . .

I was Prime-ing my system at 226 while I was at work and it locked after 2 hours, so 226 is no go. DDR450 still ain't bad, though. I'll have to see what happens tomorrow when my XMS3500 arrives. Hopefully I'll have better results to report.

 

[Duvie]

I can run my xms pc3000 at 2.8v to get 426mhz ddr stable in memtest and prime95...However epox mobo cals default 2.6v if that makes you wonder!!!!


I am disappointed to hear about the GB's so far...If I got one I could only do 170fsb with my chip anyways...

 

[Rhett]

posted in wrong thread! moved to xms shoot-out

 

[jpetermann]

We might just be waiting for a bios release for the Granite Bay boards. I am still wondering if anyone else thinks it is wierd that my GB board sees the p4m as a xeon? Maybe it is no big deal.

Thugs, Check your PM..

PJ

 

[THUGSROOK]

Originally posted by: Duvie
I can run my xms pc3000 at 2.8v to get 426mhz ddr stable in memtest and prime95...However epox mobo cals default 2.6v if that makes you wonder!!!!

makes me wonder what the real vdimm voltages are.
2.6 is really 2.6? i doubt it, it more like 2.65v probably.

anyways im talking specificly about the P4B533.
i know other boards are different

 

[Duvie]

3 different monitoring programs do usually have the epox mobo showing actual at .02v higher then what I have it set at....

 

[THUGSROOK]

> Memtest86 < all 11 tests, 10 loops
passes upto 223fsb 446ddr so far

...and this is on XMS3000

i still have to update the benchies....

 

[Mikki]

Maybe I should trade both my XMS3500 and my XMS3200 in on some XMS3000...

 

[THUGSROOK]

do you really need to ask me that at this point?

 

[Mikki]

lol!!! ...nope.

 

[THUGSROOK]

one thing thats not posted in the benchmarks ~ is how fast this system FEELS.

even at only 2400mhz 200fsb, desktop duties are instant and responsive ~ never a delay while multitasking.
its just as good if not better then my 2.26b @ 166fsb 443ddr.

 

[Mikki]

I can confirm that...was wondering if it was just me at first, glad to know you feel the same way...

 

[THUGSROOK]

retesting and repost benchmarks using XMS3000

results are WAY BETTER NOW :Q