Testing / comparing : Intel D975XBX2 / Asus P5B DX ***56K WARNING***

[shogo_ca]

commando = 965 = 6300/6400

4ghz 965 = 3.4ghz 975x

[irenic]

this board is damn good. no stability issue or overclock issue.. i'm afraid i'll regret if i buy the commando.. but to get a chance to play with a new board is very tempting haha

[sjoback]

Still have not been able to find a single program that reports the vcore voltage correctly. Core Temp reads it as 1.30, IDU reads it as 1.16 - 1.25 (changes as I'm assuming that C1E is doing it's thing) and CPU-Z reads it as 1.16??? In BIOS it's set @ 1.375 and reading as 1.325 as well.

In case you didn't find an answer, the intel desktop utilites will show proper voltages. I haven't seen anything else report proper voltages for my bx2/e6600.

[irenic]

yeah for me all the softwares from intel work well with this board (except fan control of course)

[Devious]

I just got an RMA from tigerdirect. I'm facing a similar question; get another BX2 or another board entirely (perhaps a commando).

So you still had no luck with 1.7 vMCH, I thought it fixed your problem ? Mine is good at 3.8Ghz with large FFT Orthos.

[GPSeek]

I saw you subscribe to the voodoo science of ascribing certain OC qualities to CPUs based on their release dates, which Blauhung corrected.

I did not ascribe a chip's oc-ablity to dates only. I was talking about the A/B batches. That's not voodo science. You can label any empirical obsevations/rules you do not like as voodo science.

It's a matter of luck. However, A lot of XS members here reported good results with B batch (not a stepping) chips. I would rather go with this empirical obsevation than believe the claim in above url by a so-called Intel insider. There are no such official documents or confirmed insiders. Overclockers' obsevations are much more reliable than an unconfirmed insider's claim if one is going to pick a chip.

[Ph03n1x]

yeah for me all the softwares from intel work well with this board (except fan control of course)

there is no point in changing xbx2 for commando when you have e6600...really

[theonlybabyface]

In case you didn't find an answer, the intel desktop utilites will show proper voltages. I haven't seen anything else report proper voltages for my bx2/e6600.

Yeah...that's what I'm using. The issue was with C1E. Once I disabled that, voltage was exactly where it should be. Thanks.

[uberwurst]

So you still had no luck with 1.7 vMCH, I thought it fixed your problem ? Mine is good at 3.8Ghz with large FFT Orthos.

No, could you post your settings and bios version verbatim?

1.7 vMCH enabled me to POST with > 390 FSB, but anything over 390 leads to ORTHOS/Prime95 failure, windows memtest failure, and BSODs.

My load temp is around 46-50 @ 3.495 GHz, so I certainly have the cooling headroom to squeeze out a little more juice. It's a cooperative motherboard I'm lacking

I'm probably going to just exchange.

[uberwurst]

I did not ascribe a chip's oc-ablity to dates only. I was talking about the A/B batches. That's not voodo science. You can label any empirical obsevations/rules you do not like as voodo science.

It's a matter of luck. However, A lot of XS members here reported good results with B batch (not a stepping) chips. I would rather go with this empirical obsevation than believe the claim in above url by a so-called Intel insider. There are no such official documents or confirmed insiders. Overclockers' obsevations are much more reliable than an unconfirmed insider's claim if one is going to pick a chip.

Well, this is my cpu's batch#: L631B106

I hope you are right. I'm likely going to be swapping boards and RAM pretty soon.

[Devious]

No, could you post your settings and bios version verbatim?

1.7 vMCH enabled me to POST with > 390 FSB, but anything over 390 leads to ORTHOS/Prime95 failure, windows memtest failure, and BSODs.

My load temp is around 46-50 @ 3.495 GHz, so I certainly have the cooling headroom to squeeze out a little more juice. It's a cooperative motherboard I'm lacking

I'm probably going to just exchange.

Hi, my settings were.

Disabled C1E & EIST
Enabled Enhanced power slope

Vcore 1.5 (you may need more but i dont think this is the problem)
vFSB 1.325
vMCH 1.7

Reference frequency 333
memory frequency 533 (maybe try 400 as i have to use 400 to boot at 4GHz)
tCL 4
tRCD 4
tRP tried both 4
tRASmin 11

vDimm 2.3

[GPSeek]

Well, this is my cpu's batch#: L631B106

I hope you are right. I'm likely going to be swapping boards and RAM pretty soon.

I'm pretty sure you can hit 3.60GHz = 9*400 with this board without much effort if you got decent memory sticks.
Only a few members here reported their B chips don't oc well.
So, good luck!

[heikis]

Hi, my settings were.

Disabled C1E & EIST
Enabled Enhanced power slope

Vcore 1.5 (you may need more but i dont think this is the problem)
vFSB 1.325
vMCH 1.7

Reference frequency 333
memory frequency 533 (maybe try 400 as i have to use 400 to boot at 4GHz)
tCL 4
tRCD 4
tRP tried both 4
tRASmin 11

vDimm 2.3

ref freq@ 333
that is a bit pointless, don't you think?
try if it posts at 266 ref freq.

[Blauhung]

It's a matter of luck. However, A lot of XS members here reported good results with B batch (not a stepping) chips. I would rather go with this empirical obsevation than believe the claim in above url by a so-called Intel insider. There are no such official documents or confirmed insiders. Overclockers' obsevations are much more reliable than an unconfirmed insider's claim if one is going to pick a chip.

hehe, I didn't want this to become a fighting point. Only wanted to say that the looking at the batch number won't really tell you much and it's easier to just not worry about it. I did some checking and the A/B is the revision and not the stepping. It was a terminology mixup on my part.

[GPSeek]

hehe, I didn't want this to become a fighting point. Only wanted to say that the looking at the batch number won't really tell you much and it's easier to just not worry about it. I did some checking and the A/B is the revision and not the stepping. It was a terminology mixup on my part.

No problem at all
I just want to stess that overclocking is more a matter of experience than theory. I read that kind of insiders remarks many times before on the net. It's an theory, which is unfortunately not comfimed by Intel or overclockers experience. That some specifc batches overclock better has been around for many years. I don't think any theory can easily refute these facts if Intel still wants to maintain some degree of tracability of its products.

[Devious]

Ive had the Badaxe2 and the RD600 board running this week and ran 1mb super_pi at random speeds/setting I noted them down at the time and here they are for what they are worth. The RD600 times were similar to my Asus P5W but i never got sync memory to work which is supposed to be faster, i was very surprised the badaxe 2 was slower than both.

Ive included a couple of Asus P5W results for reference.

Asus P5W

3.942 Ghz - Mem 438Mhz sync 2T 4-4-4-15-19 - 12.968s
4.000 Ghz - Mem 445Mhz sync 2T 4-4-4-15-19 - 12.797s


RD600

3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-8-4-13-20 - 13.390s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-8-4-13-20 - 13.500s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-4-4-4-20 - 13.437s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-4-4-13-8 - 13.281s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-4-4-4-8 - 13.265s

3.900 Ghz - Mem 400Mhz async 2T Strap 300 4-4-4-4-20 - 13.469s
3.900 Ghz - Mem 451Mhz async 2T Strap 300 4-4-4-4-20 - 13.343s
3.900 Ghz - Mem 500Mhz async 2T Strap 300 4-4-4-4-20 - 13.219s
3.900 Ghz - Mem 533Mhz async 2T Strap 300 4-4-4-4-20 - 13.206s
3.900 Ghz - Mem 550Mhz async 2T Strap 300 4-4-4-4-20 - 13.187

3.900 Ghz - Mem 400Mhz async 3T Strap 300 4-4-12-13-30 - 13.567s
3.900 Ghz - Mem 451Mhz async 3T Strap 300 4-8-4-13-20 - 13.328s
3.900 Ghz - Mem 451Mhz async 3T Strap 300 4-4-4-4-20 - 13.266s
3.900 Ghz - Mem 500Mhz async 3T Strap 300 4-4-4-4-20 - 13.219s

3.924Ghz - Mem 550Mhz async 2T Strap 300 4-4-4-4-20 - 13.072s



Badaxe2

3.810 Ghz - Mem 529Mhz sync 333strap 4-4-4-11-20 - 13.781s

3.870 Ghz - Mem 538Mhz sync 333strap 4-4-4-11-20 - 13.507s

3.870 Ghz - Mem 430Mhz sync 266strap 4-4-4-11-20 - 13.468s
3.870 Ghz - Mem 538Mhz sync 266strap 4-4-4-11-20 - 13.109s??


3.900 Ghz - Mem 433Mhz sync 333strap 4-4-4-8-20 - 13.546s
3.900 Ghz - Mem 433Mhz sync 333strap 4-3-3-3-6 - 13.500s
3.900 Ghz - Mem 433Mhz sync 333strap 3-4-4-8-12 - 13.531s
3.900 Ghz - Mem 433Mhz sync 333strap 3-3-3-3-6 - 13.500s
3.900 Ghz - Mem 541Mhz sync 333strap 4-4-4-8-20 - 13.438s
3.900 Ghz - Mem 541Mhz sync 333strap 4-3-3-3-6 - 13.391s

3.900 Ghz - Mem 433Mhz sync 266strap 4-4-4-8-12 - 13.375s
3.900 Ghz - Mem 433Mhz sync 266strap 4-3-3-3-6 - 13.329s
3.900 Ghz - Mem 433Mhz sync 266strap 3-4-4-8-12 - 13.359s
3.900 Ghz - Mem 433Mhz sync 266strap 3-3-3-3-6 - 13.328s

3.946 Ghz - Mem 438Mhz sync 333strap 3-4-4-15-19 - 13.375s
3.946 Ghz - Mem 438Mhz sync 266strap 3-4-4-8-19 - 13.156s

4.010 Ghz - Mem 334Mhz sync 333strap 3-4-4-15-19 - 13.328s


Super_pi run 3 times at each speed and best score taken.

[LazyBum]

Here's my best 24/7 stable settings so far with this board. Nothing earth shattering compared to some other people, but this CPU would only do 3375 completely stable on the P5W64 and on the DS3. (my old DS3 is doing 3.8 for the new owner's chip ) Using bios 2333

CPU Volts	1.525
Mem Frequency	533
tRAS	10
tCL	4
tRCD	4
tRP	4
Memory Voltage	2.24
MCH Voltage	1.600
PCI Latency Timer	32
Reference Frequency	266
FSB  Volts	1.450
FSB frequency	410
PCIE frequency	Default

It took a lot of trial and error, I pretty much built an excel spreadsheet with the settings, and upped FSB slowly, changing one thing at a time when Orthos failed until I could get stable for at least 1/2 hour. I seem to have hit a wall with this chip right here, voltage increases aren't helping anymore.

[heikis]

how does 1066strap and mem frequency 666 (divider 4:5) and 400FSB work for you guys?

[Supertim0r]

commando = 965 = 6300/6400

4ghz 965 = 3.4ghz 975x

The Commando is DAMN close or equal with Quad
Tomorrow : conroe testing and MAYBE GTL Reference Voltage on the XBX2


3DM06 3.4Ghz XBX2

3DM06 3.4Ghz Commando

[C Stat B]

Has anyone else had a problem where the latest BIOS would insist they used the 800MHz RAM divider?

[Zucker2k]

Ive had the Badaxe2 and the RD600 board running this week and ran 1mb super_pi at random speeds/setting I noted them down at the time and here they are for what they are worth. The RD600 times were similar to my Asus P5W but i never got sync memory to work which is supposed to be faster, i was very surprised the badaxe 2 was slower than both.

Ive included a couple of Asus P5W results for reference.

Asus P5W

3.942 Ghz - Mem 438Mhz sync 2T 4-4-4-15-19 - 12.968s
4.000 Ghz - Mem 445Mhz sync 2T 4-4-4-15-19 - 12.797s


RD600

3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-8-4-13-20 - 13.390s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-8-4-13-20 - 13.500s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-4-4-4-20 - 13.437s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-4-4-13-8 - 13.281s
3.900 Ghz - Mem 400Mhz async 1T Strap 300 4-4-4-4-8 - 13.265s

3.900 Ghz - Mem 400Mhz async 2T Strap 300 4-4-4-4-20 - 13.469s
3.900 Ghz - Mem 451Mhz async 2T Strap 300 4-4-4-4-20 - 13.343s
3.900 Ghz - Mem 500Mhz async 2T Strap 300 4-4-4-4-20 - 13.219s
3.900 Ghz - Mem 533Mhz async 2T Strap 300 4-4-4-4-20 - 13.206s
3.900 Ghz - Mem 550Mhz async 2T Strap 300 4-4-4-4-20 - 13.187

3.900 Ghz - Mem 400Mhz async 3T Strap 300 4-4-12-13-30 - 13.567s
3.900 Ghz - Mem 451Mhz async 3T Strap 300 4-8-4-13-20 - 13.328s
3.900 Ghz - Mem 451Mhz async 3T Strap 300 4-4-4-4-20 - 13.266s
3.900 Ghz - Mem 500Mhz async 3T Strap 300 4-4-4-4-20 - 13.219s

3.924Ghz - Mem 550Mhz async 2T Strap 300 4-4-4-4-20 - 13.072s



Badaxe2

3.810 Ghz - Mem 529Mhz sync 333strap 4-4-4-11-20 - 13.781s

3.870 Ghz - Mem 538Mhz sync 333strap 4-4-4-11-20 - 13.507s

3.870 Ghz - Mem 430Mhz sync 266strap 4-4-4-11-20 - 13.468s
3.870 Ghz - Mem 538Mhz sync 266strap 4-4-4-11-20 - 13.109s??


3.900 Ghz - Mem 433Mhz sync 333strap 4-4-4-8-20 - 13.546s
3.900 Ghz - Mem 433Mhz sync 333strap 4-3-3-3-6 - 13.500s
3.900 Ghz - Mem 433Mhz sync 333strap 3-4-4-8-12 - 13.531s
3.900 Ghz - Mem 433Mhz sync 333strap 3-3-3-3-6 - 13.500s
3.900 Ghz - Mem 541Mhz sync 333strap 4-4-4-8-20 - 13.438s
3.900 Ghz - Mem 541Mhz sync 333strap 4-3-3-3-6 - 13.391s

3.900 Ghz - Mem 433Mhz sync 266strap 4-4-4-8-12 - 13.375s
3.900 Ghz - Mem 433Mhz sync 266strap 4-3-3-3-6 - 13.329s
3.900 Ghz - Mem 433Mhz sync 266strap 3-4-4-8-12 - 13.359s
3.900 Ghz - Mem 433Mhz sync 266strap 3-3-3-3-6 - 13.328s

3.946 Ghz - Mem 438Mhz sync 333strap 3-4-4-15-19 - 13.375s
3.946 Ghz - Mem 438Mhz sync 266strap 3-4-4-8-19 - 13.156s

4.010 Ghz - Mem 334Mhz sync 333strap 3-4-4-15-19 - 13.328s


Super_pi run 3 times at each speed and best score taken.

I think my times are better than the P5W and the RD600 even though they have higher clocks. My RAM timings are a little tighter @ 4-4-4-8-19, FSB/DRAM 4:5 on the 1066 strap.

[Blacklash]

If vFSB is too low that will make 400+ unstable as well. Try 1.375v or more as needed.

You may want to keep the memory @ sloppy timings too til you figure out what is needed voltage wise elsewhere. Run 5-5-5-15 then tighten up later.

[syl]

Ive had the Badaxe2 and the RD600 board running this week and ran 1mb super_pi at random speeds/setting I noted them down at the time and here they are for what they are worth. The RD600 times were similar to my Asus P5W but i never got sync memory to work which is supposed to be faster, i was very surprised the badaxe 2 was slower than both.

I found that disabling the onboard sound improves super_pi performance significantly. Might want to give that a try

[wa2000]

Thanks gtj and DaddyRabbit.

Once I set the IDE to RAID mode on ICH7R. I can boot into XP with drives connected to the Marvell controller.

[gtj]

Thanks gtj and DaddyRabbit.

Once I set the IDE to RAID mode on ICH7R. I can boot into XP with drives connected to the Marvell controller.

Cool. Glad it worked out.