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

[Gunnz Akimbo]

Just did an OC Q6600 @3g's.
I have 4 sticks of 1GB DDR2-533 that ive upped to 1.88V.
The ram is sticks for quality, but would 1.92V kill em ya reckon?

[Woland]

Nope. You can PROBABLY go to 2.1 without any worries.

[chaotic646]

I was one of the first people to purchase this board, and mine is Rev. 503. I have been running 2X 1Gb sticks in slots 1 and 3 for about a year now with no problems, but only @ 1.9V. Im wanting to add two more 1Gb stick in slots 2 and 4 and plan on keeping all four at 1.9V. Does anyone think this will be a problem? Are there any problems with Windows XP running 4Gb of RAM? It seems like I read something a long time ago about XP not liking 4Gb with this board.

[LoneWolf]

I was one of the first people to purchase this board, and mine is Rev. 503. I have been running 2X 1Gb sticks in slots 1 and 3 for about a year now with no problems, but only @ 1.9V. Im wanting to add two more 1Gb stick in slots 2 and 4 and plan on keeping all four at 1.9V. Does anyone think this will be a problem? Are there any problems with Windows XP running 4Gb of RAM? It seems like I read something a long time ago about XP not liking 4Gb with this board.

XP runs fine with it; it just can't address all the RAM (not the 32-bit version anyway). Because of this, my install of XP reports/uses 3.25GB of RAM; I knew that when buying 4GB, but RAM was so darn cheap it made sense for down the line (I don't plan on switching to DDR3 any time soon). There are some reports of the odd application that doesn't like having more than a certain amount of RAM on a 32-bit OS, but I haven't found it yet, and I still have one 16-bit app I occasionally run left over from Windows 3.1 days that works fine.

I find that going from 2GB to 3GB of RAM has noticeable differences if you run very heavy apps (read: Adobe CS3), a large number of apps at once, or if you have a really heavy game you occasionally wish to Alt+Tab out of (e.g., Oblivion, Stalker). I'm also told that in dual-core systems, the benefit of more RAM is greater (less likelihood that an individual core will be starved for resources), but while that seems logical enough, I've done no testing of my own.

[sauria]

4 sticks of G.SKILL 4GB (4 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) F2-6400CL5D-2GBNQ 5-5-5-15 seem fine -- passes Memtest86 after hours @ stock VDIMM

[chaotic646]

Thanks for the replies. I was going to get two sticks of 512MB but two 1Gb sticks are only $13 more so it would be dumb not to even if XP can only handle 4GB. DDR2 memory is a LOT cheaper even since last year when I built my current rig. Must be because DDR3 is out now.

[Mygalito]

Hi,

I am new on this forum, and i am french. I hope you will understand me.

It is not my first quad, not my first oc, not my first memory timings settings, but it is my first Intel BadAxe2 (rev 507, bios=2792, 2797 or 2802, test with 2333 too). My job is assemblying PC for particular customers and optimize mine (photograph).

The System:
Quad Q6600 G0 @ 9x345, 9x346 but not 9x350, Crucial Ballistic Tracer (5.5.5.15, 2.16v and +, or 4.4.4.12@2.24v or 2.28, 2.32v) 266/533 or 266/667 or 333/667, with ratio 1:1 or 4:5. I change each parameter one by one. Antec P182-Seasonic II 550 Energy II

I have another quad QX6700, B3, hot on Gigabyte P35-DQ6@12x284, Crucial 4.4.3.12, any issue with it, in a beautiful Lian-Li v2000 (very large) Seasonic 600.

The cooling is by air with Thermalrigth Ultra 120 eXtreme with Artic 12025pwm or Noctua NH-12F. I prefer Artic because t° is -8° than Noctua on the same system.

My issue:

I read about the first 70 pages and i didn't found the beginning of an answer.

The Bad Axe 2 don't "consider" the voltage i set in Bios (maintenance or normal bios). I set 1.35v or 1.40v or more but the voltage is always is 1.2750 or less with speedfan, Everest or Intel DU.

C1E, EIST, Watchdog, etc are disabled. But even enable, i see the same thing.

I can set the voltage i want in bios (maintenance or normal bios). I see it, i shutdown the PC, i unplug power, i wait for green light off and i reboot, i see the good voltage in bios... but...

Under windows, i run OCCT for test 4 cores @100% during 30' and i can see the voltage idle = 1.17/1.16 (like vid) and full @1.24/1.25.

I can Never see my own voltage i set, as 1.40 or more.

I thought that my CPU did'nt need this voltage so i could'nt see this voltage.

My pc is very stable @3146MHz... but i think that the frequence of Q6600 could be set higher. Over, the cpu voltage don't go high..



Is the issue know?

Is the issue from my BadAxe2 or my Quad ?

Do you think i have to test my other quad on it ? or exchange QX6700 on BadAxe2 and Q6600 on P35-DQ6.

I test 2795, 2797 and 2802. I test too the old 2333 bios. I reset the bios each time i change....

Very strange issue...

Thanks for your explanation and the big array of frequency memory (usefull for the beginning).

[Mygalito]

I test today some others CPU, like C2D 2160.

My motherboard BadAXE2, revision 507 - bios 2802 is very strange.

It doesn' use the voltage which i set in bios over than 1.3250. Always the bios line: Current voltage = 1.3250. altough i set for example 1.40v

I think that my BadAxe2 is blocked with le voltage max of CPU, et doesn't accept others higher voltages.

So for oc the C2D 2160, it is very easy with stock voltage to obtain 3 GHz.

For Quad Q6600 G0, very easy to have 9x346 @v stock. But higher, no !!! The same voltage is used, by consequence, higher -> crash


May be i must to send it to Intel.

I spent about more than 24 effective hours with stability tests, settings...

i am tired ;-)

[chaotic646]

This is from GTJ's Bad Axe 2 guide. Hope it helps.

"The aSC7621 hardware monitoring chip also monitors 5 voltages: the standard 12v, 5v, and 3.3v supplies that do the bulk of the non-cpu work, and VCORE and VMCH (sometimes labelled simply 1.5v). The monitor is fairly accurate and will probably show the VCORE and VMCH slightly lower that set via BIOS. This is a normal artifact of digitally controlled voltage regulators. While VMCH should remain constant, the VCORE will probably change in response to power saving features like C1E and EIST. The VCORE can also change due to a phenomenon known as Vdroop. This is basically a "brownout" situation where the VRD can't supply the required voltage under extreme load. A certain amount of Vdroop is normal and, although I can't verify this, it's said that the XBX2 has one of the lowest Vdroops around.

There does not seem to be a way to monitor the memory voltage other than with a multimeter. You can do so by probing here: Memory Voltage measuring point. You can also measure VCORE and VMCH: CPU Core Voltage measuring point and MCH Core Voltage measuring point. Measure safely please. The voltages themselves aren't dangerous but an X6800 processor, for instance, can draw 90 amps. Short that to ground and the least of your problems will be getting a new board."

http://www.peakin.com/xbx2/index.html

[Mygalito]

thanks for your answer.

I know the GtJ 's page about XBX2, i read it first to understand what i have received. Nice Job, and very pleasant to read GtJ 's posts, high level !!! ;-)

I think it is the only way to be certain. But...

All the software i used, gave me the same voltage. Everest 2007 ultimate, Cpu Z 1.41, Speedfan 4.34 beta 34, Intel DU, Rightmark...

And... with my crucial, i know that it could go higher (i have another system with QX6700 with problems, so i can test each pack).

i think my BadAxe2 was modified or is special mod mother board, or limited on "going up" voltage.

I am very surprise by this.

of course, i have tested with ratio 1:1, 266/533 underclocked memory PC2-6400, set vMCH higher, set vFSB higher, set vCore > than 1.5 to test with more voltage. Always the same case, even i use Core2d or Core2quad...

so... no more idea ? RMA to Intel ?

I think i test the voltage with electronic multimeter this evening... (Not yet tested..)

thanks.

edit during the night of Paris:

I test a small but very strong e2160 (1800MHz normaly) at 3107 MHz = vcore stock, vMCH stock, vFSB stock, vDDR2 431MHz@2.28 4.4.4.12, 4:5. Well well. 9x346 like my Q6600. Over 347 or 355 or more, not stable.
The voltage delivered by BadAxe2 doesn't go over 1.3250 volts, even if i set in Bios 1.4 or more. It is an invariant of this situation.
I read all the 173 pages of this thread, and nobody has this problem.

[Salamander 2]

XP runs fine with it; it just can't address all the RAM (not the 32-bit version anyway). Because of this, my install of XP reports/uses 3.25GB of RAM; I knew that when buying 4GB, but RAM was so darn cheap it made sense for down the line (I don't plan on switching to DDR3 any time soon). There are some reports of the odd application that doesn't like having more than a certain amount of RAM on a 32-bit OS, but I haven't found it yet, and I still have one 16-bit app I occasionally run left over from Windows 3.1 days that works fine.

I find that going from 2GB to 3GB of RAM has noticeable differences if you run very heavy apps (read: Adobe CS3), a large number of apps at once, or if you have a really heavy game you occasionally wish to Alt+Tab out of (e.g., Oblivion, Stalker). I'm also told that in dual-core systems, the benefit of more RAM is greater (less likelihood that an individual core will be starved for resources), but while that seems logical enough, I've done no testing of my own.

What about 4GB of RAM with the 32-bit version of Vista - using this mobo?

[Mygalito]

In the lastest bios 2802, C1E issue is fixed. Remember, C1E enabled/disabled didn't work.

So, for my problem, i will change certainly the mother board, or i keep this wonderful stable system at "9x345@vstock-4.4.4.12-T°42 load".

[Boone.msi]

So whats the update with the 45nm chips? Still no official support? Any links to the modded support previously mentioned in this thread?

[Mygalito]

Hi,

I finished the mesurment of voltage on my BadAxe2.

So, for MCH and Memory, the settings in bios and the mesurement are equal, some different milivolts, not a problem.

I mesure vcore = 1,325 volts for 1,35 in bios, 1,475v for 1,50v in bios.

so, the mesurement shows me that the vcore is not limited "to go up" (is it good in english ?)

Thanks to chaotic646, who gave me the way to be better...

I don't understand why my oc didn' work on this board.

I 'm going to make a new oc from the beginning !!

[Lome]

This is from GTJ's Bad Axe 2 guide. Hope it helps.

"The aSC7621 hardware monitoring chip also monitors 5 voltages: the standard 12v, 5v, and 3.3v supplies that do the bulk of the non-cpu work, and VCORE and VMCH (sometimes labelled simply 1.5v). The monitor is fairly accurate and will probably show the VCORE and VMCH slightly lower that set via BIOS. This is a normal artifact of digitally controlled voltage regulators. While VMCH should remain constant, the VCORE will probably change in response to power saving features like C1E and EIST. The VCORE can also change due to a phenomenon known as Vdroop. This is basically a "brownout" situation where the VRD can't supply the required voltage under extreme load. A certain amount of Vdroop is normal and, although I can't verify this, it's said that the XBX2 has one of the lowest Vdroops around.

There does not seem to be a way to monitor the memory voltage other than with a multimeter. You can do so by probing here: Memory Voltage measuring point. You can also measure VCORE and VMCH: CPU Core Voltage measuring point and MCH Core Voltage measuring point. Measure safely please. The voltages themselves aren't dangerous but an X6800 processor, for instance, can draw 90 amps. Short that to ground and the least of your problems will be getting a new board."

http://www.peakin.com/xbx2/index.html

FYI droop is done on purpose in order to prevent the opposite of droop known as overshoot

[silverfox2]

Suddenly the computer freeze when booting up. It stops with E9 and will not go further.
I can access bios but all seems okey, any suggestion ?

[Comp-Freak]

Dunno if the code displayed is the same as the code on port 80h, but if so, E9 means entering BIOS setup according to the (online) manual. Bit weird if you didn't push F2. You can try to boot in maintenance mode.

[silverfox2]

I have no problem entering Bios with F2. Everything seems ok in bios, no problem whatsoever.
The problem is when I boot the computer it freeze with E9 on screen.

You cant boot in maintenance mode, only get acess to other options " watchdog " and other things...

[wa2000]

I have no problem entering Bios with F2. Everything seems ok in bios, no problem whatsoever.
The problem is when I boot the computer it freeze with E9 on screen.

You cant boot in maintenance mode, only get acess to other options " watchdog " and other things...

Try with only keyboard and monitor connected to the computer, see if it boots ok.....

[silverfox2]

Try with only keyboard and monitor connected to the computer, see if it boots ok.....

Still the same problem..

[wa2000]

Still the same problem..

Are you using USB keyboard? If yes, try with a PS/2 keyboard....

Or try to take the whole system apart, then try to boot it with just CPU, 1 stick of memory and the display card.

[silverfox2]

Are you using USB keyboard? If yes, try with a PS/2 keyboard....

Or try to take the whole system apart, then try to boot it with just CPU, 1 stick of memory and the display card.

I am using a PS/2 keyboard, and have already tried to boot with only Cpu, memory and displaycard.
I have also tried my friends Cpu, memory and displaycard..
And the problem still remains...

I think I have to RMA this board, thanks for trying to help.

[screwtech02]

Anyone running the newer 3850/70's in this board??? I'm curious as to the performance enhancments/improvments to be made/had???

[edjrwinnt]

I just realize that I have the older B3 "step" version of the Q6600. I'm only stable at about 3 GHZ. Anyone get much higher than this with this version of this chip?

[joe.denice]

hey guys, i was wondering since I am using 667 ram and I have a Q6600 and want to OC it to 3ghz which is a 333 fsb, what should I set the ram settings to be?? Thanks for the heads up.