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

[GPSeek]

On the first HS there was on all edges a little bit rough but nothing in the middle of the HS.

On the new one I don´t know because the HS is about 3hours on it I will check this in a few days.

I think your IHS is concave. Mine is concave too. You need either put more AS5 at the center or lap the IHS

Another factor is your cpu batch number. Those B batch CPUs are hot.

[pushook]

thx for the info^^ i will try on weekend with more grease in the middle.
what do you mean with the B?
On my intel box reads"BX80557E6600
FPO# L627B042
Version#63625-002

[GPSeek]

thx for the info^^ i will try on weekend with more grease in the middle.
what do you mean with the B?
On my intel box reads"BX80557E6600
FPO# L627B042
Version#63625-002

L627B042

You only scored a B. That's better than an A in real life.

[uberwurst]

I just tried reseting to default -> retweaking my 2935 bios, and I my system still BSODs at 3.6

1.425v cpu, 1.4 fsb, 1.675 mch; This time my RAM wasn't failing at all, at 5-5-5-15 2.2v (at a frequency of exactly 400), and it was memtest stable for about 20 mins.

I also disabled C1E, which did not stop the constant BSODs.

I can hardly open the start menu before I get a BSOD. I think I'm gonna try 2333.

Right now everything is OK @ 3.495, which seems to be a wall I'm hitting. This is annoying

[pcdoc1]

I just tried reseting to default -> retweaking my 2935 bios, and I my system still BSODs at 3.6

1.425v cpu, 1.4 fsb, 1.675 mch; This time my RAM wasn't failing at all, at 5-5-5-15 2.2v (at a frequency of exactly 400), and it was memtest stable for about 20 mins.

I also disabled C1E, which did not stop the constant BSODs.

I can hardly open the start menu before I get a BSOD. I think I'm gonna try 2333.

Right now everything is OK @ 3.495, which seems to be a wall I'm hitting. This is annoying

It's a 2.4 processor, why would believe it should run 3.6 at any vcore... Some do, most don't.

[Suge Knight]

I did not see any difference when the C1E is disabled. I always have as much problem with the motherboard, it boot well at 444 MHz FSB and then of a blow, it refuses to boot, it is very irritating.

Curiously , the chart is faster into 4/4/4/12 than into 3/3/3/8:

4/4/4/12:




It's not optimised ^^

[pushook]

I also reflashed from 2395 to 2333 today and now the problems with the keyboard has gone. And no more reboots in 3Dmark05^^ Also the bug with the ram has gone for me!

[chaotic646]

L627B042

You only scored a B. That's better than an A in real life.

Sorry if I'm misunderstanding something but my 6600 is an A. Are the B batch supposedly better?

[wa2000]

After I flashed to 2395 using Intel's stupid recommended fashion (the .exe way), I had no problems for hours until I shut it down and rebooted the next day and it wouldn't POST. Only the fans/lights/mobo LEDs would power on, but the fans would never cycle down below 100% speed and I'd never get a successful POST regardless of how many times I swapped jumper positions, cleared CMOS, etc. It's just a flawed BIOS. Really made me just want to beat something.

If your board would not post, stick your ear close to the speaker/buzzer of the board. If you hear couple beeps(very low volume) when booting up, it can be recover.

First remove the power cord or turn off the PSU. Ensure the small yellow light close to the ram slots is not on.

Then put the bios jumper into position 2-3. Then remove the CMOS battery for 30 seconds then put the battery back in.

Connect the power or turn the PSU back on. Turn on the computer, now you should able to go to the BIOS config screen.

Everytime after I did the above steps, only 2 values are incorrect in BIOS. Date/time and memory speed and voltage settings. Memory setting always went back to AUTO.

[Supertim0r]

Sorry if I'm misunderstanding something but my 6600 is an A. Are the B batch supposedly better?

yes, they tend to clock better

[GPSeek]

Sorry if I'm misunderstanding something but my 6600 is an A. Are the B batch supposedly better?

supposedly better overclocking if you read carefully here.
However, it's generally agreed B batch is hotter than A.

[Omastar]

IMO, B chips OC slightly better than A on the whole, but I'd rather have significantly lower temps than an extra 100MHz.

[chaotic646]

yes, they tend to clock better

Figures that I got an A then. I have a few degrees to spare but I think I'm maxxed out @ 400 FSB.

[Suge Knight]

Little test with 2.8V and Crucial PC8000:


I succeeded to boot at 500 Mhz cas 3 (fsb = 333 Mhz) but the systeme crash when i open CpuZ. I hope we will have quickly a soft to overckock in windows to do maxscreen.

[screwtech02]

Any update on the "next" bios release????

[Blauhung]

supposedly better overclocking if you read carefully here.
However, it's generally agreed B batch is hotter than A.

hehe, i'll go into my rant on batch numbers that I've tried to impart before

the A/B is actually stepping, in CPU-z you can also see revision #'s. Everytime the stepping/rev changes that means some lithography mask gets reissued. The bulk of those changes don't actually end up with any change to the actual transistor layout or metal layers, but are more to improve manufactureability (not a word, but I don't really care). This is probably the only place where you will show up with any minor change to the ability of the chip.

Beyond that you end up with what week they were produced in. A lot of people like thinking that a given week of packaging means that you will be much more likely to have a good overclocker. Along the line, a lot will probably take anywhere from 20-40 weeks to make it from bare silicon to final cut die in its package. Over that time any number of process changes (most of which are put in place to cut down on cost and improve production tool use) could be made over one of the thousands of steps that the wafers go through. Then all the die on the wafers get tested and binned. After that they get shipped overseas to one of the assembly plants where they get cut up and soldered to a board and have a IHS put on. Lots can be waiting for 1-10 weeks just sitting around waiting for their turn to get hacked into. So the actual week that a chip gets packaged has very little to do with when that chip worked its way through the factory.

Some also feel that the factory that the chip ends up getting packaged at has a bit to do with its ability. If anything the factory that the wafer gets produced at has way more of a pull on how the chip will clock. they all just get sent out to the place that has room for the next shipment. So the factory that the wafers get cut at has very little to do with how the chip will do.

Beyond that you also have to factor in where your chip was in the lot that it moved through the factory in. conroe is 25 wafers per lot, something like 350 die per wafer. Each wafer in a lot ends up slightly different due to process variations that depend on location in the batch. Each die in a wafer ends up slightly different due to the location on the wafer.

More then likely tracking the week that a chip's batch (which doesn't even directly relate to what lot the wafer originally occupied because batches are assigned after the dies are tested, sorted, and binned) has little affect on the true overclockability of a chip. With all the variation that a chip goes through how well it does when you start testing it has almost nothing to do with the serial number on the front.

hehe. sorry guys, sometimes coming from the inside and seeing people clamoring over what week a chip is just kinda funny to me. Although you were only talking about stepping here, i didn't want it to start wandering down that road [/rant]

[uberwurst]

No matter what I do, including a rollback to 2333, tame RAM settings, raised vcore, disabled C1E, among many other tactics, I can't get mine past 3495 GHz with any stability. And I don't mean ORTHOs unstable, I mean BSOD unstable.

cpu: L631b106 , mobo: rev 505

3495 is certainly good enough, although I wonder what makes my conditions so much different from everyone else's.

FUNNY how I selected most of my parts before seeing this thread, and my system is almost identical to Supertim0r's, right down to the tuniq tower, teamgroup RAM and OCZ gamexstreem PSU. UNFUNNY how fragile my system is compared to his ;c

[Omastar]

I think you just got a bit of a crappy chip. Week 30 and week 32 both > week 31.

[Blauhung]

I think you just got a bit of a crappy chip. Week 30 and week 32 both > week 31.

sigh...

[lad]

Teratism wrote: I do not really know how, but the board has a jumper setting for fixing a bad bios with a floppy. You take the jumper off, shove in a recovery diskette and it is supposed to fix it.



I can verify that this works...

[uberwurst]

sigh...

lol.... it could be that I got a lame chip, but judging it by its week is voodoo science man... listen to Blauhung.

[lad]

I'm running 2395 with an E6600 at 3.5 GHz (385 FSB). Just went 10 hours on Orthos blend...

Ran 2 hours on small FFT's last nite, and an hour with StressTest on everything except printer and DVD tonight (so far) while ripping CD's and reading XS.

[Big Worm]

$189.98 from mwave.com

http://www.mwave.com/mwave/viewspec....iteria=BA23276

[nealh]

$189.98 from mwave.com

http://www.mwave.com/mwave/viewspec....iteria=BA23276

nice price..wonder if its rev. 505

[wa2000]

2 975XBX2 mb, one board takes ~20-30 seconds before showing the Intel logo screen after pressing the power button. The other one just takes ~10 seconds to show the Intel logo. Both board have BIOS 2395.

Also the 2 boards have different readings in voltage and CPU temperature in IDU.

Just switch the mb and everything else remain the same.