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

[gtj]

Links working now...

New Fixes/Features:
 Updated Microsoft Vista* support for STAC9227 and STAC9274D Audio
Codecs.
 Fixed boot order for SCSI HDD with certain SCSI controllers.
 Adding Serial Port enable/disable setup question in Standard BIOS.
 Added programming of PCI Subsystem Vendor ID and Subsystem ID for
the PCIe Graphics bridge.
 Fixed issue where system would not boot after installing a Vista*
OEM Activation string.
 Fixed issue where certain host bus adapter card was not
functional.
 Added code to program the OEM SVID and SSID for onboard Marvell*
controller.
 Reduced S3 resume time related to processor detection.
 Fixed issue where 64-bit operating system installs would blue
screen with certain processors.

[gtj]

I dont want to fiddle right now to test this out, but do you think it will bomb out your system if you are running higher voltage to get a higher freq? Is it quick enough to actually work as advertised?

I have C1E and EIST turned on on 2 of my systems now and it's not giving me any problems.

[gtj]

Here is what I accomplished so far:

I set the:
Reference Frequency to 266
Memory Frequency 533
I left the memory voltage at default, which was 1.84V

CPU:
Enhanced Powerslope enabled
CPU voltage override 1.375V
Host Clock Freq: 377

also disabled watchdog, C1E, VT, XD, EIST

so far I'm stable at 3.40GHZ with 52C temp. Orthos ran stable for more than 10hrs.

but now, when I try to get to 3.6, the system crashes.
And here is another dilemma I have : my memory is rated at 4-4-3-5 with voltage 2.0-2.1, but in the CPUz it shows 5-5-5-15. Is that normal? I still can't figure out why it changes by itself, and still don't know if 4-4-3-5 is better.

thank you guys

I think you're going to have to up your cpu voltage. Go right to 1.5 and see if it's stable, then back it down if you can.

Memory manufacturers program the SPD very conservatively so the sticks will boot ofter install. 5-5-5-15 is typical for PC6400 memory. You'll have to change it in the BIOS to get better timings. If you go to a 400 FSB you'll be running the memory at it's rated speed but lowering the timings will probably require you to raise the voltage. If it's rated at 2.0 - 2.1, try 2.08 to start.

[chaotic646]

I just installed the newest bios and for some reason my e6600 is running at 1.6 instead of 2.4. I loaded up the default bios before I updated them and it set it successfully back to 2.4. After the update its now running at 1.6. Very strange...

[horatiub]

I think you're going to have to up your cpu voltage. Go right to 1.5 and see if it's stable, then back it down if you can.

Memory manufacturers program the SPD very conservatively so the sticks will boot ofter install. 5-5-5-15 is typical for PC6400 memory. You'll have to change it in the BIOS to get better timings. If you go to a 400 FSB you'll be running the memory at it's rated speed but lowering the timings will probably require you to raise the voltage. If it's rated at 2.0 - 2.1, try 2.08 to start.

gtj, yeah, my goal will be to run stable at 3.6 so if I can go to 400FSB would be nice. So, basically, you're telling me to go into the bios and change the memory timings to the original ones, 4-4-4-3, up the voltage a little and run tests.

Also, isn't it 1.5v a little high for the CPU?

[chaotic646]

I just installed the newest bios and for some reason my e6600 is running at 1.6 instead of 2.4. I loaded up the default bios before I updated them and it set it successfully back to 2.4. After the update its now running at 1.6. Very strange...

EDIT: Okay my Bios is still showing 2.4 but cpu-z is showing 1.6 with a multiplier of 6 and when I go to contol panel\system it shows 1.6.

[theonlybabyface]

Are both EIST and C1E disabled?

[chaotic646]

Are both EIST and C1E disabled?

No, but C1E is for voltage not cpu speed. Everything is default so an e6600 should by default run at 2.4. I think its a problem with the new bios but cant confirm just yet.

[theonlybabyface]

No, but C1E is for voltage not cpu speed. Everything is default so an e6600 should by default run at 2.4. I think its a problem with the new bios but cant confirm just yet.

I know but my cpu speed would drop as well unless I had both disabled. It was really weird...

[supa-fly]

Ok guys, all links to Intel new bios files working.

Go go go...

[gtj]

EDIT: Okay my Bios is still showing 2.4 but cpu-z is showing 1.6 with a multiplier of 6 and when I go to contol panel\system it shows 1.6.

That's definitely EIST. Even when disabled, Everest will tell me it's running at x6 then it'll jump to x9 when I put it under load.

I think it's a BIOS bug.

[gtj]

gtj, yeah, my goal will be to run stable at 3.6 so if I can go to 400FSB would be nice. So, basically, you're telling me to go into the bios and change the memory timings to the original ones, 4-4-4-3, up the voltage a little and run tests.

Also, isn't it 1.5v a little high for the CPU?

You shouldn't have any problem being stable at 400fsb. Start by leaving the memory alone at 5-5-5-15 1:1 266/533 then when you get a stable 400fsb lower the timings. 1.5v is just a starting point. The point is to get a stable 400fsb before monkeying around with other stuff. So get stable, then start changing things one at a time keeping good records of what you've tried.

[horatiub]

You shouldn't have any problem being stable at 400fsb. Start by leaving the memory alone at 5-5-5-15 1:1 266/533 then when you get a stable 400fsb lower the timings. 1.5v is just a starting point. The point is to get a stable 400fsb before monkeying around with other stuff. So get stable, then start changing things one at a time keeping good records of what you've tried.

gtj, thank you.

so now a million dollar question: am I going to kill my CPU if I let it overclocked for years? It might sound stupid, but I always thought about it

also, as far as I know, when I run 1:1 266/533, I'm on the 1066 bootstrap, so I'm assuming that I won't lose any performance

[gtj]

gtj, thank you.

so now a million dollar question: am I going to kill my CPU if I let it overclocked for years? It might sound stupid, but I always thought about it

also, as far as I know, when I run 1:1 266/533, I'm on the 1066 bootstrap, so I'm assuming that I won't lose any performance

Doesn't sound stupid at all. The short answer is "yes". Running the CPU with a higher voltage and the corresponding higher temperatures will shorten the life of the CPU. Whether there's any emperical evidence to say exactly how much, I'm not sure. I haven't seen any but I haven't looked all that hard either. Maybe it's 8 years instead of 10, I don't know. Anyway, an E6600 at 3.6g/400fsb, shouldn't have it's life materially shortened as long as you use the lowest stable voltage and keep the temperatures within reason. If you're going to be running the CPU flat out 24x7, something better than air cooling may be required.

266/533 is the 1066 strap so you're good there.

[irenic]

anyone here knows any xeon processor that can be used with this xbx2 ? thx

[chaotic646]

You can mix and match to your heart's content. Both controllers will support mixes of drives in an array and standalone drives and the BIOS will support booting from any of them.

From your description, your best bet is to create a RAID 0 on your Intel controller with the 2 new drives. That'll get you the best performance.

In BIOS, set the Intel controller mode to RAID and reboot. You'll get a new Intel Matrix BIOS prompt where you can press Ctrl-I and create your array. Reboot again and go back to the BIOS boot order menu and make sure the marvell array is still set first in the boot order. Boot to Windows and install the Intel drivers and you should be good to go.

gtj - I just got around to trying this but it doesnt seem to work. I set the Intel controller to RAID, I hit ctrl-i and created my array, I booted to Windows from the marvell array, but I cant load the intel matrix drivers from the floppy. The readme.txt says you have to load them by hitting F6 during the OS install, which I knew but I dont want to install my OS on my new Intel array. Is there any way I can install the drivers from windows on my marvell array or am I going to have to install windows on my new array too just to load the drivers?

EDIT: Nevermind, I feel dumb. I just realized I could load them off of the Intel Express Driver CD. Its working fine now.

[topboy]

OK Here's the deal.

I was running all A OK on bios 2333 with Vista 64 bit with Crossfire enabled on the latest 7.1 Catalyst driver.

I then upgraded the bios to Intels latest 2618 and boom. The bios stops the cards from running in Crossfire mode.

The Catalyst Control Centre kept hanging, and putting errors in the Applications event log of windows.

I attempted 4 uninstall's and reinstalls of the catalyst drivers with no luck, same errors.

Logged a call with ATI and they are stumped.

I then thought mmm rollback the bios, surely not, it couldnt be the bios. OH YEAH it was.

Rolled back to 2333, uninstalled the Radeon Catalyst drivers, rebooted and reinstalled. Rebooted and all is back in good working order. Drivers all OK and Crossfire is working.

This is on a brand new install of Vista on freshly formatted drives today so i know its most definatly a problem with this latest 2618 bios from Intel.

This is also with default values in the bios, NO Overclock.

Just thought i would share this all with you

Topboy

[theonlybabyface]

anyone here knows any xeon processor that can be used with this xbx2 ? thx

Yup...any 775 will work. Mine works great!!!

[theonlybabyface]

OK Here's the deal.

I was running all A OK on bios 2333 with Vista 64 bit with Crossfire enabled on the latest 7.1 Catalyst driver.

I then upgraded the bios to Intels latest 2618 and boom. The bios stops the cards from running in Crossfire mode.

The Catalyst Control Centre kept hanging, and putting errors in the Applications event log of windows.

I attempted 4 uninstall's and reinstalls of the catalyst drivers with no luck, same errors.

Logged a call with ATI and they are stumped.

I then thought mmm rollback the bios, surely not, it couldnt be the bios. OH YEAH it was.

Rolled back to 2333, uninstalled the Radeon Catalyst drivers, rebooted and reinstalled. Rebooted and all is back in good working order. Drivers all OK and Crossfire is working.

This is on a brand new install of Vista on freshly formatted drives today so i know its most definatly a problem with this latest 2618 bios from Intel.

This is also with default values in the bios, NO Overclock.

Just thought i would share this all with you

Topboy

Thanks. That's a reason for me to stay away. Any others experience the same issue???

[topboy]

Can someone who lives in the USA or Canada call Intel on 916-377-7000 and report this problem please as they will only talk to US or Canadian Residents and the UK office is now closed.

I think this is a pretty serious problem with this latest bios and they should know about it.

Cheers

Topboy

[gtj]

OK Here's the deal.

I was running all A OK on bios 2333 with Vista 64 bit with Crossfire enabled on the latest 7.1 Catalyst driver.

I then upgraded the bios to Intels latest 2618 and boom. The bios stops the cards from running in Crossfire mode.

The Catalyst Control Centre kept hanging, and putting errors in the Applications event log of windows.

I attempted 4 uninstall's and reinstalls of the catalyst drivers with no luck, same errors.

Logged a call with ATI and they are stumped.

I then thought mmm rollback the bios, surely not, it couldnt be the bios. OH YEAH it was.

Rolled back to 2333, uninstalled the Radeon Catalyst drivers, rebooted and reinstalled. Rebooted and all is back in good working order. Drivers all OK and Crossfire is working.

This is on a brand new install of Vista on freshly formatted drives today so i know its most definatly a problem with this latest 2618 bios from Intel.

This is also with default values in the bios, NO Overclock.

Just thought i would share this all with you

Topboy

I had an issue with my nvidia cards as well although it resolved itself. Basically Vista redetected both the video card and the marvell raid controller and reinstalled the drivers for them. Had me scared for a bit but after another reboot, everything came back.

[topboy]

Yeah i think from the readme of the bios they have re-written the PCI stuff and have broken it at the same time.
Quote from the readme "Added programming of PCI Subsystem Vendor ID and Subsystem ID for the PCIe Graphics bridge."

This would explain why your video cards and the raid controller got redetected in Vista.

Looks like Intel messed up with this latest bios in certain situations, i.e. crossfire.

[gtj]

Yeah i think from the readme of the bios they have re-written the PCI stuff and have broken it at the same time.
Quote from the readme "Added programming of PCI Subsystem Vendor ID and Subsystem ID for the PCIe Graphics bridge."

This would explain why your video cards and the raid controller got redetected in Vista.

Looks like Intel messed up with this latest bios in certain situations, i.e. crossfire.

I can't step out this meeting (boring as it is) to make a call but I just submitted a support request.
http://supportmail.intel.com/scripts...d=36,2386,2578

You should do the same thing at least to get some activity going.

[topboy]

I can't step out this meeting (boring as it is) to make a call but I just submitted a support request.
http://supportmail.intel.com/scripts...d=36,2386,2578

You should do the same thing at least to get some activity going.

Done thanks for the link.

Is anyone else able to verify this problem. If someone else has ATI and Crossfire can you upgrade your bios to 2618 and try reinstalling the 7.1 catalyst drivers. You can then rollback to another bios after.

[screwtech02]

Done thanks for the link.

Is anyone else able to verify this problem. If someone else has ATI and Crossfire can you upgrade your bios to 2618 and try reinstalling the 7.1 catalyst drivers. You can then rollback to another bios after.

I'll try it when i get home from work, i'm runnin XP-64, so will that matter???