New P5W DH (lot:6CM) + Kentsfield

[kyosen]

I've just got a new P5W DH.
Its lot is 6CM, so manufactured in Dec. 2006, I think.


The board revision is 1.04G...
...Vmch=1.85V setting is available, of course.


Then I tried FSB overclockability test with Kentsfield QX6700.
The test is very simple...finding max FSB for ORTHOS 3min running.
Test condition: Vmch~1.85V and MCH water cooled for each case.

Results:

P5W DH(lot:67M) ~ 342MHz
http://www.oohashi.jp/c-board/file/O..._342.5_3-4.png

D975XBX2(REV.502) ~ 384MHz
http://www.oohashi.jp/c-board/file/O...XBX2_383.9.png

P5W DH(lot:6CM) ~ 398MHz


The lot of MCH chip is:
My old P5W DH (lot:67M) -> L552A722
My D975XBX2 (REV.502) -> 3625B181
My new P5W DH(lot:6CM) -> 3644B182

So, newer MCH brings more overclockablity for Kentsfield, it seems.
Yes I know that P5B series(P965) are more Kentsfield friendly,
but 975X is still slightly faster than P965 at same clock...
...P5W DH is not dead, especially for multiplier unlocked Kentsfield

[^don.k's^]

And what about Conroe FSB OCability ? Have you tried ? I'm a p5w lover and i'd like to see more than 500fsb... ( to dream is for free )



PS: i mean without vmods, i get 500 @2,15vmch

[TaPaKaH]

Yes, dual-core OC we also want to see

[eva2000]

oooooooooh looks like they're nearly getting there... maybe squeeze out 450fsb on 975x for kentsfield will suit most Q6600/QX6700 users

[death metal]

Very nice results kyosen, thanks for sharing...

[kyosen]

Today I tried X6800 running on this P5W DH(6CM)...
...hmm...I'm very confused
In 4:5 mode, memory clock can exceed 480MHz easily (1GBx2, CL3-3-3-4),
but in 1:1 mode, max FSB is around ~450MHz, so far.

...and now, I must go home...I'll return in several days

[eva2000]

Today I tried X6800 running on this P5W DH(6CM)...
...hmm...I'm very confused
In 4:5 mode, memory clock can exceed 480MHz easily (1GBx2, CL3-3-3-4),
but in 1:1 mode, max FSB is around ~450MHz, so far.

...and now, I must go home...I'll return in several days

tRD, and tRFC related probably ?

[Gautam]

One thing I noticed is that with my ram, CAS 3 is actually more stable than CAS 4 1:1... At CAS 4 I got stuck at ~455-460, at CAS 3, 475.

[DJSUB]

any new overclocking updates on 1.04g revision?

tnx

[nicepun]

Today I tried X6800 running on this P5W DH(6CM)...
...hmm...I'm very confused
In 4:5 mode, memory clock can exceed 480MHz easily (1GBx2, CL3-3-3-4),
but in 1:1 mode, max FSB is around ~450MHz, so far.

...and now, I must go home...I'll return in several days

It's the FSB, not the memory.

One thing I noticed is that with my ram, CAS 3 is actually more stable than CAS 4 1:1... At CAS 4 I got stuck at ~455-460, at CAS 3, 475.

Now is that CAS 3 475 1:1? It's the FSB on this mobo, more than the memory itself.

[toxo]

Quote:Originally Posted by TWY

I see.. Wonder if it has any hardware enhancements for Kentsfield, etc. Intel XBX1 and XBX2 had some changes in traces, wonder if it's the same for the 1.03G vs. 1.04G on the P5W.

IIRC the newer 975X chipsets are more rectangular in shape (only seen where the factory/year/week code is 36**), whereas the older ones are square (I've seen L5**). If I'm not wrong the rectangular ones can clock better with lower vMCH? The 975XBX2 should be using the rectangular ones. Could anyone confirm this?

http://resources.vr-zone.com/Shamino/badx2/20.jpg


XBX1's chipset vs. the XBX2's chipset

How i said in the other post:

I´m new here, but i´m reading yours for a long time. My P5W is a 1.04G rev, 1.85v MCH and BX2. I´m using 1901. Is fine for me. No problems yet. My config is 400x8 1:1 with 6400 L629A817 with 1.35 in bios and 1.55v MCH. Stable in orthos and OCCT. I´m using 4 sticks KVR ECC with Promos IC´s.

Conclusion: is 1.04g rev beter for oc with new chipset rev??

[TWY]

How i said in the other post:

I´m new here, but i´m reading yours for a long time. My P5W is a 1.04G rev, 1.85v MCH and BX2. I´m using 1901. Is fine for me. No problems yet. My config is 400x8 1:1 with 6400 L629A817 with 1.35 in bios and 1.55v MCH. Stable in orthos and OCCT. I´m using 4 sticks KVR ECC with Promos IC´s.

Conclusion: is 1.04g rev beter for oc with new chipset rev??

Looks like it clocks better..

My 1.03G needs 1.65V MCH for stable 400x9 1:1 with 6600 L630A. Things get worse when I start to use dividers. I need to bump MCH to 1.65V at 360x9 if I use 4:5. 360x9 1:1 is still stable with 1.55V MCH. Using the Crucial 10th Anniversary 2GB kit.

[jozzy]

One thing I noticed is that with my ram, CAS 3 is actually more stable than CAS 4 1:1... At CAS 4 I got stuck at ~455-460, at CAS 3, 475.

what vdimm are you used to hit 475 CAS 3 ?? nice fsb

[Leeghoofd]

my P5W is in the RMA since today , hope to get a new batch version then that would be nice ....

[TWY]

Trying to get my hands on one 1.04G...

[BulldogPO]

What version number has that better 975X chip?

[Gautam]

what vdimm are you used to hit 475 CAS 3 ?? nice fsb

Board's now dead (unrelated) but it was 2.8v vdimm and 2.2v mch or so. vdimm wasn't wholly necessary.

[Lane-k]

Looks like it clocks better..

My 1.03G needs 1.65V MCH for stable 400x9 1:1 with 6600 L630A. Things get worse when I start to use dividers. I need to bump MCH to 1.65V at 360x9 if I use 4:5. 360x9 1:1 is still stable with 1.55V MCH. Using the Crucial 10th Anniversary 2GB kit.

Same prob with E6700.. impossible to boot at 4:5 after 380x10.. boot and setting TRD to 7 with memset, fix the prob for me.. then i can hit 420x10.. In real the memory need this TRD 7 for pass 500mhz (DDR not FSB..).. Like i can't set it in bios i need put it manually... board is 1.04G, manufactured 7 dec 06 (if i don't do an error reading the lot..)

After read Kyosen, i will test FSB/memory speed properly.. Before i got my RMA P5W64