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

[Lome]

Actually C1E *does* decrease the vcore when it switches down to the low-speed multiplier, which is why the cpu draws less power and runs cooler...

yeah but only when the pwm controller isnt locked down due to manual set voltage
most of the power savings are from entire parts of the processor being shut down and this is not a feature of c1e just normal c1 -e

[qckings]

Need some help here:

My goal is to get a 1Ghz Overclock on my E6600 chip (Batch number L631F577). I am close to getting it but am running into Orthos failures. This is the message I get and it happens within 5 seconds of the test:

FATAL ERROR: Rounding was 0.5, expected less than 0.4 Hardware failure detected, consult stress.txt file.

My current settings which are stable with Prime 95 and Orthos are:
3.32 OC
FSB 370
MEM 533 Timings are original 4-4-4-12 @ 2.0v GSkill DDR-800
Ref Feq 266
CPU voltage 1.3250
FSB voltage 1.250
MCH 1.525
My temps are 39/56C
My case is a CMStacker 830 with 6 120 MM fans and a 9700 LED fan on CPU.

First, I just bump the FSB to 380 and left everything else the same. Orthos fails after 5 seconds or so. Then I raised the CPU vcore up to 1.350 and Orthos failed. Then I raised the MEM voltage to 2.12 and Orthos still failed. What am I not doing right? Any suggestions? Thanks in advance.

[scorpion]

qcking,you need to raise your vcore ,vmch, vfsb, a nocth .

[scorpion]

make sure the watchdog is disabled

[nealh]

yeah but only when the pwm controller isnt locked down due to manual set voltage
most of the power savings are from entire parts of the processor being shut down and this is not a feature of c1e just normal c1 -e

well look at my post and that is what I said...blah

[qckings]

scorpion,

Watchdog is disabled. I just forgot to mention it on my initial post. I had the vcore and the MEM upped two notches already. I will up the vmch and vfsb and see what it does. I noticed in your sig that you did not raise the MEM voltage any, should I lower it back to 2.00?

qcking

[qckings]

here is an update to upping the voltages:

raised the vcore from 1.375 to 1.4125
raised vfsb from 1.300 to 1.325
raised vmch from 1.525 to 1.550
mem voltage is 2.08v

Ran Orthos and within 10 minutes it failed. My temps were hitting 64C during test.

[scorpion]

no mem voltage i myself would keep aroud 2.00 to 2.04 try up your vcore to 1.50.what is your fsb.

[scorpion]

vcore 1.475 to 1.50

[qckings]

I will bump up the vcore to 1.50. My FSB is 380.

[scorpion]

there was a misprint vdimm is 2.04

[scorpion]

qckings, once you get it stable try droping your vcore down a nocth.

[ubiquityman]

Looks like the S3 suspend problem with the BX2 is not fully resolved yet.

Intel sent me a new mobo, rev 507. (previously had 505).

With the latest 2747 BIOS, there is still a problem coming out of suspend with SATA drives connected to the Marvell (crap) SATA controller.

Basically the drives that are connected to the Marvell controller often disappear after resuming from suspend.

Sometimes a "Scan for hardware changes" in devices manager will fix the problem. Other times, I have to Disable the Marvell controller in device manager then re-enable it.

Really unfortunately that Intel put this Marvell controller on the BX2.

On an unrelated note, I realized that the motherboard audio can be configured to drive a front panel MIC. The pinouts are not in the manual (at least not in the version that I have), but they are on the Intel web site. There is a BIOS setting for selecting "Legacy Audio" vs "HD Audio".

[qckings]

scorpion,

I did what you mentioned above and could not get Orthos to run more than 7 minutes. My temps hit 43/71C with Speed Fan. My vcore was 1.500 and then I lowered it to 1.475 and both times could not run Orthos. The rest of my settings were:

FSB 380
MEM 533 Timings are original 4-4-4-12 @ 2.00v
Ref Feq 266
CPU voltage 1.500 then 1.475
FSB voltage 1.325
MCH 1.550

[Woland]

Might need 1.55vcore.

[ArcticOC]

New BIOS out!

Download Version: 2770

BIOS Version 2770
About This BIOS:
 June 13, 2007
 BX97520J.86A.2770.2007.0613.1403
 SATA RAID info:
 Intel(R) RAID for SATA - v5.6.2.1002
 Marvell 88SE6145 SATA BIOS – v1.1.0.34
 PXE UNDI: Initializing Intel(R) Boot Agent GE v1.2.28
 PXE Base Code: Intel(R) Boot Agent GE v1.2.42
 Memory reference code equivalent to MRC 1.2+
 Intel® Active Management Technology info:
 Intel® AMT BX Firmware: v1.2.3
New Fixes/Features:
 Updated processor support.
 Fixed resource allocation issues with PCI cards with multiple P2P
bridges.

[rbsmontemor]

New BIOS out!

Version: 2770

BIOS Version 2770
About This BIOS:
 June 13, 2007
 BX97520J.86A.2770.2007.0613.1403
 SATA RAID info:
 Intel(R) RAID for SATA - v5.6.2.1002
 Marvell 88SE6145 SATA BIOS – v1.1.0.34
 PXE UNDI: Initializing Intel(R) Boot Agent GE v1.2.28
 PXE Base Code: Intel(R) Boot Agent GE v1.2.42
 Memory reference code equivalent to MRC 1.2+
 Intel® Active Management Technology info:
 Intel® AMT BX Firmware: v1.2.3
New Fixes/Features:
 Updated processor support.
 Fixed resource allocation issues with PCI cards with multiple P2P
bridges.

I'm going to test NOW

Rbs.

[Xilikon]

Curious to know what added cpu support ?? added 1333 FSB support ??

[ChaosMinionX]

Curious to know what added cpu support ?? added 1333 FSB support ??

1333FSB needs a hardmod to the board to support it. Check out the thread on Penryn

Here you go





source:http://anandtech.com/cpuchipsets/int...spx?i=2972&p=2

The mod is under neath the 3rd blue heatsink, which isnt currently present on production BX2

[aicjofs]

1333FSB needs a hardmod to the board to support it. Check out the thread on Penryn

Here you go





source:http://anandtech.com/cpuchipsets/int...spx?i=2972&p=2

The mod is under neath the 3rd blue heatsink, which isnt currently present on production BX2

Are you sure you need a hard mod for 1333FSB? I believe all you need is BIOS support for 1333FSB. It's voltage options for 45nm penryn that I believe you need the hardmod for, hence why it's under a VRM heatsink, probably a resistor to allow lower voltage.

[Xilikon]

Are you sure you need a hard mod for 1333FSB? I believe all you need is BIOS support for 1333FSB. It's voltage options for 45nm penryn that I believe you need the hardmod for, hence why it's under a VRM heatsink, probably a resistor to allow lower voltage.

Yeah, IIRC, the new Penryn stuff require both a 1333 fsb support and some new VRM features. I believe it may be possible to run but without the new VRM features.

[scorpion]

qckings, try what woland said, some cpu take more vcore than others.have you try runnning everest.

[scorpion]

you might want to bump up your vdimm a notch.

[scorpion]

sorry about such a slow post been out the loop.

[qckings]

scorpion,

I am afraid to bump anything higher because my temps are over 71C right now. What puzzles me is I can have a FSB of 370 which puts me @ 3.34 Ghz with everything at original voltages STABLE and as soon as I go beyond 370 the computer is very UNSTABLE and then a lot of extra voltage is required to get to 3.4Ghz. I am willing to go as high as 65C load. Maybe water cooling is the way to go.