Official ASUS X38 Boards: P5E, P5E3 - Info/BIOS Thread

[STaRGaZeR]

P5E BIOS 501 is out! ftp://ftp.asus.com.tw/pub/ASUS/mb/socket775/P5E/

Who is so brave to test it?

Not me, after the last one I´m done with BIOS updates with this board for now

[Brahmzy]

0402 is still rock solid here...not changing unless I hear great additional benefits....

[WeldZilla]

I am still working with 107 and have found no need to update from it yet.

WZ

[TKC]

OK, so I have flashed 0501, for everyone here in the forum , following is my finding by now

1. bug fixed, bois setting can't change value by "+" & "-" button unless number is inputted in advance.
2. new bug, divider 4:5 AND 5:6 do not post when FSB > 400.
3. bug still exist, CPU Q-fan do not work. (Asus please wake up !)
4. bug still exist, report inaccurate temp and intermittently stop reporting.
5. to be confirmed, fsb ceiling around 470~480.

overall experience isn't good for 0501, I am forced to decrease fsb due to divider problem, you may stay where you are, and I am waiting for the next update.

[Brahmzy]

OK, so I have flashed 0501, for everyone here in the forum , following is my finding by now

1. bug fixed, bois setting can't change value by "+" & "-" button unless number is inputted in advance.
2. new bug, divider 4:5 AND 5:6 do not post when FSB > 400.
3. bug still exist, CPU Q-fan do not work. (Asus please wake up !)
4. bug still exist, report inaccurate temp and intermittently stop reporting.
5. to be confirmed, fsb ceiling around 470~480.

overall experience isn't good for 0501, I am forced to decrease fsb due to divider problem, you may stay where you are, and I am waiting for the next update.

Thanks for the detailed report TKC!

[ea6gka]

the tragedy continues , afaict

[Brama]

OK, so I have flashed 0501, for everyone here in the forum , following is my finding by now


2. new bug, divider 4:5 AND 5:6 do not post when FSB > 400.

Not true in my rig.

I am 430x8, strap 333 and 5:6 and it is stable like a rock!

With my hardware and settings, it is by sure better than 402.

[TKC]

Not true in my rig.

I am 430x8, strap 333 and 5:6 and it is stable like a rock!

With my hardware and settings, it is by sure better than 402.

Maybe this issue is hardware dependent, but it's out of question that 0501 is much worse than 0107 in my board, FSB/DRAM 435x8/1088(0107) vs 400x8/1066(0501).

I also found that 0501 have problem on SpeedStep(EIST), EIST will stop working after some period of ocing, and the option of speedstep in the bios will disappear! WTF.....

Everyone please keep feedback so we can discover how new bios play, as Asus never give out clear information about their bios.

BR

[WeldZilla]

you bet ea6kga I see you are no further with your new board than you were with you P5E. I on the otherhand am loving this new torture tool from Asus. Them disgusting bastards have forced me to just keep pushing this POS farther and tighter everyday!! Boy they are just completely disgusting dawgs.
I mean How am I to live with this new board when I can only get it to run stable with my Q6600 at a lousy 3862mhz. with my mem on the 4:5 and these crappy 4-4-4-4 timings Oh the misery I live with daily. I guess I am just a sucker for punishment.
Jeez I guess I should RMA this POS.

I mean the damn thing only sets a new high for itself everyday since I got it.

Check out this poor lame new high cpu score for 06.

http://service.futuremark.com/compare?3dm06=4136743

I am going to write Asus a long letter and really tell them off!!!

WZ

[ea6gka]

you bet ea6kga I see you are no further with your new board than you were with you P5E. I on the otherhand am loving this new torture tool from Asus. Them disgusting bastards have forced me to just keep pushing this POS farther and tighter everyday!! Boy they are just completely disgusting dawgs.
I mean How am I to live with this new board when I can only get it to run stable with my Q6600 at a lousy 3862mhz. with my mem on the 4:5 and these crappy 4-4-4-4 timings Oh the misery I live with daily. I guess I am just a sucker for punishment.
Jeez I guess I should RMA this POS.

I mean the damn thing only sets a new high for itself everyday since I got it.

Check out this poor lame new high cpu score for 06.

http://service.futuremark.com/compare?3dm06=4136743

I am going to write Asus a long letter and really tell them off!!!

WZ

jesus , that is a monster clock

[TKC]

With crippled 0501 bios, I still manage to 3.6G finally.
Updated to my signature.

[sauria]

With crippled 0501 bios, I still manage to 3.6G finally.
Updated to my signature.

Did you do the VDroop mod?

[Amurtigress]

Hello,

I could use a few opinions on needed FSB voltages for the FSB clock area of 400 MHz and above on the P5E.

I've had intermittent crash problems on my P5E and on my former P5B Deluxe board. Only yesterday I managed to get my system rocksteady. Anyway, my main problems were sudden application crashes, like Second Life and Need for Speed Underground and Carbon. The frequency of the crashes varied widely from every few seconds (Vista 64, Need for Speed Undergound 2), up to every few days (Second Life in Win2003 Server x86).

I came to the conclusion that my FSB voltage was too high, and the NB core volt too low at times. What makes my system stable currently is:

CPU divider: 8x (E6600 at 3.6 GHz, 1.46V)
NB voltage: 1.55V
FSB clock: 450 MHz
FSB voltage: 1.26V
FSB Strap: 400
RAM: 4x1 GB Corsair XMS-2 PC800 CL5 at 2.1V/900 MHz

What are your experiences for the best FSB and NB voltages?

PS: Orthos has always been stable for me after 6h continuous stress tests in blend mode.

Looking forward to your replies.

[Brahmzy]

Hello,

I could use a few opinions on needed FSB voltages for the FSB clock area of 400 MHz and above on the P5E.

I've had intermittent crash problems on my P5E and on my former P5B Deluxe board. Only yesterday I managed to get my system rocksteady. Anyway, my main problems were sudden application crashes, like Second Life and Need for Speed Underground and Carbon. The frequency of the crashes varied widely from every few seconds (Vista 64, Need for Speed Undergound 2), up to every few days (Second Life in Win2003 Server x86).

I came to the conclusion that my FSB voltage was too high, and the NB core volt too low at times. What makes my system stable currently is:

CPU divider: 8x (E6600 at 3.6 GHz, 1.46V)
NB voltage: 1.55V
FSB clock: 450 MHz
FSB voltage: 1.26V
FSB Strap: 400
RAM: 4x1 GB Corsair XMS-2 PC800 CL5 at 2.1V/900 MHz

What are your experiences for the best FSB and NB voltages?

PS: Orthos has always been stable for me after 6h continuous stress tests in blend mode.

Looking forward to your replies.

Similar to mine - I've got an E6600 running 425x8 (3.4Ghz) @ 1.425V. Mem's at 5:6 1066 2.1V.
Are you sure you need that much NB Voltage? I'm rock solid at 1.35 NB V. Maybe that extra 25 FSB needs a lot more...

What week is your E6600? Sounds like a very similar chip to mine - I need about 1.45V-1.46V to do 3.6Ghz.

[Amurtigress]

Similar to mine - I've got an E6600 running 425x8 (3.4Ghz) @ 1.425V. Mem's at 5:6 1066 2.1V.
Are you sure you need that much NB Voltage? I'm rock solid at 1.35 NB V. Maybe that extra 25 FSB needs a lot more...

What week is your E6600? Sounds like a very similar chip to mine - I need about 1.45V-1.46V to do 3.6Ghz.

No, I'm not sure about the NB voltage. I'll set it to 1.30 or 1.35 for a test tonight.

My week code is 7638A742 (Yes, no trailing L), and SSPEC is SL9S8

I just realized that my RAM is getting a little unstable at 900 MHz, so I am back to 875 (RAM) and 3.5 GHz CPU (437x8).

I wonder what's better. 400x9/RAM at 800 or 437x8 at 875 RAM clock. I doubt tho it really matters...memory speed doesn't make much of a difference on Intel systems compared to AMD64s.

[mascaras]

Q6600
444x9
Aircooled @ thermalright extreme (with scythe Ultra-Kaze 120mm @ 2000 rpm)
Asus P5E (bios 0107)
Strap = 333
NB voltage = 1.47v (didnt try less)
Vcore = 1.55v (didnt try less)

[Amurtigress]

Q6600
444x9
Aircooled @ thermalright extreme (with scythe Ultra-Kaze 120mm @ 2000 rpm)
Asus P5E (bios 0107)
Strap = 333
NB voltage = 1.47v (didnt try less)
Vcore = 1.55v (didnt try less)

That basically answers my question, Strap 333 requiring more voltage. Cool overclock, tho!

Has anybody observed application crashes upton too high FSB voltages? Namely , 1.38V and more on Strap 400 and 333?

I have had several instabilities for months, and I seem to have fixed all of them at some point during that time but not all of them. I only see now what has been wrong, which is hard to tell apart into separate problems anyway when all of them are causing the same issues on the P5B Deluxe. It finally made me buy the P5E, which solved them by giving me the Strap 400 now.

Which is the main reason why I wanted it.

BTW, the NB is now running at 1.31V, stable. Have a look at my specs, I finally got my 8800GTS to a whopping 660 MHz.

---
ASUS P5E, BIOS 0402
E6600 @ 3.5 GHz/1.46V/437*8
X38 MCH volt: 1.31, Strap 400, FSB 1.26V
4 GB Corsair XMS2 DDR2-800
Sparkle 8800GTS 320MB, GPU: 660 MHz, RAM: 2050 MHz, 3DMark 06: 11200@1280
Enermax Liberty 400W PSU
ST3500630AS SATA2/NCQ 500GB
WD2500KS SATA2 250GB

[death]

Q6600
444x9
Aircooled @ thermalright extreme (with scythe Ultra-Kaze 120mm @ 2000 rpm)
Asus P5E (bios 0107)
Strap = 333
NB voltage = 1.47v (didnt try less)
Vcore = 1.55v (didnt try less)




and thats on air what is the comp out side or some thing

wtf is the comp out side

[TKC]

Did you do the VDroop mod?

No, I didn't, and I won't do it unless Vcore/load can't support my CPU ocing anymore, even though VDrop is that large.

If VDrop can be resolved as easily as 2B pencil drawing, why mobo maker still allow it exist after so many years? In fact some opinion that I heard believe there are techincal reasons behind the VDrop.

Regards

[askimo]

Generally speaking, what is the best bios version to work with this mobo?

[Amurtigress]

No, I didn't, and I won't do it unless Vcore/load can't support my CPU ocing anymore, even though VDrop is that large.

If VDrop can be resolved as easily as 2B pencil drawing, why mobo maker still allow it exist after so many years? In fact some opinion that I heard believe there are techincal reasons behind the VDrop.

Well, Intel specifies that drop in the specifications for the voltage regulations. Often some VRM specs version like 10.5 or 11 is mentioned in mainboard feature lists...there is a long history behind those VRMs. VRM means Voltage Regulator Module.

Back int he times of Pentium 1 and Pentium Pro systems and servers, some CPUs came with a dedicated voltage module for them to provide the right voltage. This was short-lived but the modules still exist, but now integrated on the boards.

However I cannot tell you the actual reasons why the dropping was introduced later.

Hope that helps. But whenever you hear someone whining "My board is crap, the voltage drops. Bwahaha, mommie!" just blame it on a bigmouthy kiddie with no experience behind the keyboard.

Besides, the regulation chips require external capacitors. And capacitors are as they are, you can simply draw a graphite line over them. It is as it is. The circuitry outside the chips is always similar.

Why go to lengths to prevent people from doing pencil mods and still put overclocker options into your BIOS?

[Brahmzy]

Generally speaking, what is the best bios version to work with this mobo?

There is no answer for that. Everyone here seems to be using a different BIOS version because that is what seemed to worked fro them. I'm using 0402, wihich has been great for me, but others hate it, vice-versa.

Kinda strange that way.

[TKC]

askimo,

I have used 0107, 0402, 0405, and 0501, my opinion is the best one goes to 0107.


Amurtigress,

Thanks for the information.
As pencil mod to minimize Vdrop is indeed an interesting trick, someday I may use it w/o hesitation at all, but not now .

[eyes.only]

Tried again today to get some extra mhz out of my q6600 with my p5e.
Was tunning on 3200mhz(8X400) with 1,3875 vcore in bios, rest more or less default.

Used some settings for northbridge and fsb voltages and upped the speed to 9x400 = 3600 an had my vcore to 1.55 in bios to get this setup to run. Started with memtest86 and it was stable for about 7 hours. In windows vista a different story: occt running and vista crashed with a blue screen saying something about an 0X00000124 error. Had my vcore to 1.55 in bios to get this setup to run. Changed some settings but kept keeping these errors.

Any ideas?

Edit:
Vista boot with 2 gigs of ram is also not very fast. And small tricks for that?

Thanks!

[askimo]

There is no answer for that. Everyone here seems to be using a different BIOS version because that is what seemed to worked fro them. I'm using 0402, wihich has been great for me, but others hate it, vice-versa.

Kinda strange that way.

Yes very strange, perhaps the problem resides with the kind of memory used.