680i SLI is too expensive? Asus as the solution : P5N32-E SLI PLUS

[Radical_53]

Sure thing I'm still undecided whether I should try one of the newer 780i offerings. I guess it wouldn't make a big difference for now.
I'm still too curious why the FSB is so limited for quads. Last week I ran some tests with a chiller, 8°C cold water, and was able go to close to 4.2GHz. The board is capable of all this... "normally"

[panfist]

I think I may have fried this board or something with my e8400-

I had it overclocked to 3.6ghz stable for a while with stock cooling on the NB and SB. Then one day it just reboots....ok, it must be overheating a little. Except it wouldn't reboot. Not after cooling down for 1 min, not after CMOS reset, not after 20 mins.

I'm beginning to get a little worried and finally I try one more time and it works. Currently it's stable at 3.0Ghz.

When it boots up, it takes extra long for all the fans to spin up, and when it finally starts to POST, the fans lag for a bit as it starts, like the voltages all sag. I'm worried to push the system any further so I have returned all the components to stock speeds.

I wonder if I screwed up the power supply? It's an Antec SP-500.
http://www.antec.com/specs/SP500_spe.html

[Jultsu]

I got to 465fsb (with 0803), could boot to Windows but it would freeze immediately when running prime95.

After that I flashed 0901 back in again, just out of curiosity. I'll try going 450+ fsb with this one.



-----------------------------------------------------------

well, 0901 was as horrible as i thought it would be. I flashed 0801 in and now I'm working my way up from 333fsb, trying to get it stable at every step. Now I'm stuck at 362fsb. It seems to be a rollercoaster ride with the VTT voltage (around 355-370 fsb at least)

[onthepulse]

Hi all, I need some help as I am hitting a bit of a wall between 3-3.1GHz.

Does anyone have an E2180 on this board? If so, could you please send me all of the BIOS settings required to achieve an OC > 3.2GHz?

Also, can someone please explain what the LDT multiplier (defaults to 5x) does or how to use it? Is this like Hypertransport on the AMD?

=============================================

Here is what I have tried...
- Stable at 3GHz (9x333) at 1.4V with only minimal changes to other voltages. Also stable at 8x375, but not at 10x300.
- I have raised the NB voltage a little, to 1.35 and the VDIMM to 1.925 - as I am running 4 sticks of RAM.
- Seem to have some success with 'linked', but my RAM will stop me from OCing too much, so really need to use 'unlinked' to get to higher FSB
- Going to 3.1GHz, I need to raise the Vcore and also VTT (1.45). But can't get stable.
- I have tried raising all of the voltages one by one and doesn't seem to make a huge difference.
- Raising the Vcore gets me more time in Orthos before failure, but can't get beyond about 10mins
- As I am on stock cooling, I don't want to raise the Vcore much beyond 1.45. Have pushed it to 1.4875, but that doesn't help much with stability
- Might try 0803 or 0901 BIOS to see if this helps...

I feel like there is something I am missing? Or perhaps I just need better cooling to get further? My feeling is that I should be able to go higher even with the stock cooling, but at a point, I will need better cooling to keep the temps from reaching dangerous levels.

[Tw1st3d]

Don't upgraded to 0901 bios.. I stopped me from gettin a bit stabel.. So i downgraded to 0801 and that did help...

But i'm also confused with overclocking this board and E2160 SLA8Z M0

You need disable all spread spectrum and put the LDT to ?

[Radical_53]

@jultsu: That's quite a difference there. I'm still using this bios 901 because of my sound card and my RAM. Right now it's ok for me as I have that open multiplier, I might need to change it too when I get back to my dual core.
You're right about that rollercoaster. Every 50MHz increase on the CPU forced me to play around with all the voltages for quite a while to get it stable

@onthepulse: My advice would be to check FSB and RAM first, so a lowered multiplier might help for the first part of testing. Also, you could try to put as much cooling in as possible for a test. For example, if you put a large 120mm fan below your graphics card and maybe an 80mm fan on top of it, see if that helps at all to gain some stability.
This board can clock pretty well, but it can be very painful to get it there Voltages too low affect stability just like when they're set to high, especially for the board voltages. Keep all those setting on something other than auto, otherwise the board will increase them on its own.

[onthepulse]

Hi again,

I can POST at 3.2 with mild voltages. Problem is, that once I raise the Vcore above 1.45V, it stops POSTing. I have tried upping pretty much all of the voltages with no luck. Any suggestions? Could it be the stock cooling? Or is there some value that you need to adjust to get the Vcore over 1.45 and stable?

I have established that 355x9 is relatively stable. 355FSB seems to work fine at most multis, and I get best results with the 9x multi. Just can't seem to POST with enough voltage to make it Orthos stable. Getting some nice bech results though

[Tw1st3d]

Hmm i got my E2160 running now with 400 fsb, but Orthos crashed 1 sec after i started it.. And some times i get BSOD / reboots..

Vcore: 1.475
Mem: 2.0
1.2V: 1.4
NB: 1.4
SB: 1.5
CPU VTT: 1.4

PCI locks = 100 mhz
SPP = 200 mhz

LDT = x5

All spread spectrum and others are disabled

Memory is running 800 mhz@stock@5-5-5-15 T2 unlinked

So i have tried to raise volt on a bit up and down.. didn't really changed much

[onthepulse]

Hmm i got my E2160 running now with 400 fsb, but Orthos crashed 1 sec after i started it.. And some times i get BSOD / reboots..

Vcore: 1.475
Mem: 2.0
1.2V: 1.4
NB: 1.4
SB: 1.5
CPU VTT: 1.4

PCI locks = 100 mhz
SPP = 200 mhz

LDT = x5

All spread spectrum and others are disabled

Memory is running 800 mhz@stock@5-5-5-15 T2 unlinked

So i have tried to raise volt on a bit up and down.. didn't really changed much

Exactly where I'm at Anyone got suggestions for more stability?

[Tw1st3d]

Exactly where I'm at Anyone got suggestions for more stability?

I cranked the volt up to 1.575 o_O resulted in 30 sec stabil ORTHOS and then reboot...

I'm confused, but the E2160 just up and wants extreme of volt when you get over 3.4 ghz

70 celcius load temp with water o_O with the 1.575 volts

[Jultsu]

Hmm i got my E2160 running now with 400 fsb, but Orthos crashed 1 sec after i started it.. And some times i get BSOD / reboots..

Vcore: 1.475
Mem: 2.0
1.2V: 1.4
NB: 1.4
SB: 1.5
CPU VTT: 1.4

PCI locks = 100 mhz
SPP = 200 mhz

LDT = x5

All spread spectrum and others are disabled

Memory is running 800 mhz@stock@5-5-5-15 T2 unlinked

So i have tried to raise volt on a bit up and down.. didn't really changed much

NB 1.6v
1.2v HT 1.45v
SB 1.65v


try every VTT voltage setting to see which one works the best (doesn't freeze etc.)

[Jultsu]

A NEW BIOS!! WOHOHOOO!

ftp://dlsvr02.asus.com/pub/ASUS/mb/s...LIPlus1002.zip

Wonder what it does? ;D

[Radical_53]

Yeeehaa! No changelog? Ok, but it sounds interesting to try...

[Jultsu]

And now it's gone from the server :O what?

Well, I flashed it in and got BIG problems. Couldn't even boot to windows 9x333, it would give this strange "hardware malfunction" BSOD which i've never seen before. Upping the VTT to 1.55v seemed to do the trick, I'll try more O/C:ing on sunday.

----------------------------------------------------------------

And, here's my mirror if you are brave enough to try:

http://warez.pp.fi/1002.BIN


----------------------------------------------------------------

This 1002 seems to be more "forgiving" in terms of VTT voltage. I just rammed my way to 9x400, NB 1.65v and VTT 1.55v. I can't remember it being this easy with 0801. It's not stable yet though, but I can type this message from windows, so I'm halfway there already

[>GSXR<mrbusa]

clocks will not stick at anything but auto for me with any muti. i can change them in the bios but at boot the chip is at stock. also i have to break out a foppy drive because it will not let me revert to any older version with ez flash. wow afudos here i come

[Jultsu]

you mean with 1002 version? I had that same problem also, but eventually they were saved. Try changing the voltages first, then boot, then back to bios and change the fsb and save-exit.

[>GSXR<mrbusa]

you mean with 1002 version? I had that same problem also, but eventually they were saved. Try changing the voltages first, then boot, then back to bios and change the fsb and save-exit.

yes, 1002 version
i did.
i cant even get the ratio to save to anything but auto.
i am having probs flashing to an older version also because between the dos files, awdflash, and the bios file it ends up being to much data for a floppy. i have never used a cd for flashing or a usb stick other than with ezflash.

[onthepulse]

This is a VERY BAD BIOS. I now can't flash back to previous versions! I managed to save my settings once, but now it always defaults back to stock.

Really don't know how to get around this. NOT HAPPY

EDIT: Only way to change the clocks seems to be to re-flash and then on the 2nd boot, I can save. Hope this gets worked out.
I can save settings now, after re-flashing, but can't flash back to ANY previous versions.

[>GSXR<mrbusa]

you lucked out man...
i have reflashed three timed with the same result.
what even pisses me off more is the fact that in the manual it says we can flash from a floppy with awdflash but all the files needed will not fit on a floppy. i spent a few hours trying to build three versions of a boot cd, M$ way, barts pe way with his precompiled clean boot cd, and an image of two floppies with nero's floppy emulation boot cd maker.
if there are any hard core coders who could figure out how to build a workable image to fix our problem then you would have a friend for life. from what i can tell there are some bytes within some of the dos boot files that needs to be changed in order to get awdflash to work from a boot cd burned from the floppy image.
it is really going to sux if asus takes there usual couple months to fix a bios.

[onthepulse]

you lucked out man...
i have reflashed three timed with the same result.
what even pisses me off more is the fact that in the manual it says we can flash from a floppy with awdflash but all the files needed will not fit on a floppy. i spent a few hours trying to build three versions of a boot cd, M$ way, barts pe way with his precompiled clean boot cd, and an image of two floppies with nero's floppy emulation boot cd maker.
if there are any hard core coders who could figure out how to build a workable image to fix our problem then you would have a friend for life. from what i can tell there are some bytes within some of the dos boot files that needs to be changed in order to get awdflash to work from a boot cd burned from the floppy image.
it is really going to sux if asus takes there usual couple months to fix a bios.

You can put BIOS files on a USB stick and Flash from there using EZFlash utiltiy in the BIOS itself. That's how I have been doing it.

But I can't flash back to a previous version from 1002. It says something like "Downgraded ROM not suitable..."

[>GSXR<mrbusa]

You can put BIOS files on a USB stick and Flash from there using EZFlash utiltiy in the BIOS itself. That's how I have been doing it.

But I can't flash back to a previous version from 1002. It says something like "Downgraded ROM not suitable..."

Exactly...
i have always used EZflash and have never had any issues. I get the same error you do when trying to revert to an older bios from this 1002 version. my settings will not save no matter how many times i reflash the 1002 version. you will find that even the asus resource cd will not work, nor will any methods for making a bootable cd or floppy to restore an older bios version. the only thing going to save these boards if you flashed this bios is a new bios from asus that will force a flash. does anybody know if there is a syntax command for awdflash to force a flash? if there is then the asus resource cd might work to get the shipping bios version flashed.

[Jultsu]

Hmm. Did you guys really try it long enough? I found that on the fifth or sixth time my fsb settings WERE actually saved. The first four or five times they just went back to AUTO.

btw. try setting the voltages to other than AUTO, disable all spread spectrums and all other unnecessary crap, and set the SPP-MCP manually to 200MHz. Do these even stay in memory, or do they revert back to auto?

Anyways, if you do manage to save the settings, this BIOS seems to be quite good, actually. I've never ran 9x400 before this easily, I think I can get it even stable.

[>GSXR<mrbusa]

well i have been running it since i seen this post yesterday right after you posted the find of the bios. i have reflashed it four times now and my clocks will not save.

i figured out that there are only three files needed from a dos boot floppy to boot so i was able to make a working flash floppy that everything fit on.
EZ flash will not downgrade and now i was able to use awdflash and it says the same thing. awdflash will not let me force the flash either, gives a disk not ready error.

all the settings but my clock speeds stick as they should, and i have always set everything possible with manual settings so i understand what you are saying. if the clock speeds would set then i would be content for now.

[Jultsu]

Yea, well, I really don't have any other suggestions, than:

go to bios, set fsb/mem, save and exit, go to bios, set fsb/mem, save and exit, go to bios, set fsb/mem, save and exit, go to bios, set fsb/mem, save and exit, go to bios, set fsb/mem, save and exit, ......

Worked for me (linked). Did you try unlinked only?

[onthepulse]

Here's what's happening for me now... Pretty good OC results, but some strange behaviour...

- I can change FSB/MEM easily.
- I have it set on 'unlinked' and that 'sticks' 2nd boot after flashing
- I am having trouble setting the multi to anything other than 10x (default) and making that stick
- That being said, I have never had success with the 10x mutil with other BIOSes, but on this one, it works well
- My OC results are actually very pleasing I can get to 3.1GHz (10x310) - which was my previous best at any rate, and it's quite stable.
- One thing I can suggest if you want to change multi, is knock your clocks down first, then save, then change on next boot. Seems to be the only way.

So, a mixed bag... Would like the ability to go back to a previous BIOS, but at the same time, my results are really good.

Wonder if this BIOS works with Wolfdale/Yorkfield?