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

Quote:

---

Originally Posted by Jodiuh

Gotta love our board heaters, eh? :p: How are you measuring the MB Temp in Windows? I'm fairly certain Speedfan will do it, but is there an Intel tool? Also, I picked up an Antec 2 slot blower and on low, it's under the noise floor of my Scythe SFLEX E @ 1200RPM. @ med it's perhaps a touch louder, but still quieter than the 6 120mm fans in the stacker. :D

---

I've used Speedfan and Everest, mainly because they're two of the only utilities I've found that even work in Vista 64. Using Everest right now while on the internet, and in my apartment with a 24c ambient I'm actually reading 26c for my 2 cores, 39 idle for my MB, and 62 for the 8800GTX, which is the only heat culprit that I have. The MB would be at least 8c cooler with a less powerful card.

Every fan is on low and 120mm, so it's almost silent. Using the spot cool to blow 8800 air out the two slots next to the card, also on low, which has dropped the video card temps a little. My CPU core readings are insanely low. Guess a front intake, Ninja fan, and 2 exhaust fans do their job. :)

Been debating whether or not to order the Antec 2-slot blower. Would certainly be more efficient than the spot cool as it would run right next to the 8800 vents. Only concern was noise, which you say is fine. Think I'll get one cheap off amazon.

[QUOTE=gtj]1.5v isn't a problem provided you can keep your temps in line. I run 1.55v.

QUOTE]

ugh. crash-fest. i cant get 3.3ghz. tried 3.2ghz @ 1.475v. windows boots fine. orthos crashed after 15mins.

brought everything back to 3.0ghz @ 1.45v. now orthos wont run for more than 10secs. same exact settings i have been using all week.
ugh.

--
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 1789 Time: 15
CPU: 3006MHz FSB: 266MHz [261MHz x 11.5 est.]
3/7/2007 8:51 PM
Launching 2 threads...
1:Using CPU #0
2:Using CPU #1
1:Beginning a continuous self-test to check your computer.
1:Press Stop to end this test.
2:Beginning a continuous self-test to check your computer.
2:Press Stop to end this test.
1:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
2:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
2:FATAL ERROR: Rounding was 0.5, expected less than 0.4
2:Hardware failure detected, consult stress.txt file.
2:Torture Test ran 0 minutes 9 seconds - 1 errors, 0 warnings.
2:Execution halted.

1:Torture Test ran 0 minutes 9 seconds - 0 errors, 0 warnings.
1:Execution halted.
---
then windows was crashing, w/o doing anything (again at same settings as before = 3.0)
im back to stock... try orthos now.

update: at stock.. orthos wont run for more than 5mins. same error
Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4

i set my memory according to OCZ suggestions
4,4,4,15 @2.08v

thoughts?? im kicking myself for messing this up now.

update: now im getting all sorts of odd errors in windows... flickers.....
this isnt looking good.

QUOTE]
i set my memory according to OCZ suggestions
4,4,4,15 @2.08v

thoughts?? im kicking myself for messing this up now.
[/QUOTE]

I have noticed that the board resets the memory dividers to stock after each failed overclock attempt. The result is that when you crank the fsb back up without readjusting the memory strap/freq your memory ends up being massively overclocked.
Not sure if that's your problem but I was tearing my hair out over it before I figured out what it was doing.

[QUOTE=argylesocks]

Quote:

---

Originally Posted by gtj

1.5v isn't a problem provided you can keep your temps in line. I run 1.55v.

QUOTE]

ugh. crash-fest. i cant get 3.3ghz. tried 3.2ghz @ 1.475v. windows boots fine. orthos crashed after 15mins.

brought everything back to 3.0ghz @ 1.45v. now orthos wont run for more than 10secs. same exact settings i have been using all week.
ugh.

--
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 1789 Time: 15
CPU: 3006MHz FSB: 266MHz [261MHz x 11.5 est.]
3/7/2007 8:51 PM
Launching 2 threads...
1:Using CPU #0
2:Using CPU #1
1:Beginning a continuous self-test to check your computer.
1:Press Stop to end this test.
2:Beginning a continuous self-test to check your computer.
2:Press Stop to end this test.
1:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
2:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
2:FATAL ERROR: Rounding was 0.5, expected less than 0.4
2:Hardware failure detected, consult stress.txt file.
2:Torture Test ran 0 minutes 9 seconds - 1 errors, 0 warnings.
2:Execution halted.

1:Torture Test ran 0 minutes 9 seconds - 0 errors, 0 warnings.
1:Execution halted.
---
then windows was crashing, w/o doing anything (again at same settings as before = 3.0)
im back to stock... try orthos now.

update: at stock.. orthos wont run for more than 5mins. same error
Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4

i set my memory according to OCZ suggestions
4,4,4,15 @2.08v

thoughts?? im kicking myself for messing this up now.

update: now im getting all sorts of odd errors in windows... flickers.....
this isnt looking good.

---

Make sure you've completely removed power from the motherboard whenever you make a voltage change. I.E. Pull the plug and wait for the little green LED on the MB to go out. If you don't the VRM doesn't pick up the new voltage.

[QUOTE=gtj]

Quote:

---

Originally Posted by argylesocks

Make sure you've completely removed power from the motherboard whenever you make a voltage change. I.E. Pull the plug and wait for the little green LED on the MB to go out. If you don't the VRM doesn't pick up the new voltage.

---

hmm. so shut down, pull the plug, power on, make changes, reboot?

anyways... something is really messed up.
completely stock, all settings (ram at factory settings)
ref 266 mem 533

wont run orthos for more than 5 seconds
windows is now unbareably slow.

i did a system restore... then things seemed ok, until i tried orthos. again, immediate BSOD, crash.

i wish i could turn the clock back 4hours.

im at a loss of what to do.
my cpu ussage is tweetering between 50-100% while using IE.
random windows errors keep popping up.
could i have corrupted my Vista instal?
Could it be bad RAM? how do i test that?

Quote:

---

Originally Posted by argylesocks

hmm. so shut down, pull the plug, power on, make changes, reboot?

anyways... something is really messed up.
completely stock, all settings (ram at factory settings)
ref 266 mem 533

wont run orthos for more than 5 seconds
windows is now unbareably slow.

i did a system restore... then things seemed ok, until i tried orthos. again, immediate BSOD, crash.

i wish i could turn the clock back 4hours.

im at a loss of what to do.
my cpu ussage is tweetering between 50-100% while using IE.
random windows errors keep popping up.
could i have corrupted my Vista instal?
Could it be bad RAM? how do i test that?

---

Hang in there.

Make your changes THEN completely power down the machine. When you change voltage settings in BIOS, they don't take effect until you remove power and restore it.

So, recreate your stable environment in BIOS, then do the powerdown and reboot.

EDIT: Don't start making hardware changes unless it was a hardware change that triggered the problem!

Quote:

---

Actually 266/667 is 4/5 divider. CPU-z reports the same. Shows Mem freq @ 462.4 currently. Unless both gtj's handy memory calc page and CPU-Z are both wrong it's 4:5. I see 266/533 as 1:1 on both as that is what I was running for my OCing testing. Found CPU limit and then started to see which memory speed was stable. So far this is stable. I was thinking of trying 200/533 on the 800 strap to get some more bandwidth out of it and better memory performance on the lower strap.

---

Yes, yes...sorry...I was mistaken. Stupid error. 266/667 4/5, of course. Good, anyway, i think 2.3v is too much. Of course is stable, but beyond 2.1v in many Mobos (680i mainly) the memories are being fried after a month of use. It is a subject that is becoming a great controversy

@argyle
Investigate what processes are burning 50-100% cpu with task manager.
You can test memory right from the boot screen in vista. I've only used Vista in dual boot and one of the boot options is windows memory tester. You may have to use f8 right after getting out of bios to see this option if you're single booting it as it may not pause at that screen.

if you F8 a single boot Vista installation, you get a special boot options screen for vga mode / safe mode / force non digitally signed drivers etc.....

if you then hit esc to cancel, it forces the multiboot options screen and gives you the windows memory diagnostic option....

Quote:

---

Originally Posted by stmckin

if you F8 a single boot Vista installation, you get a special boot options screen for vga mode / safe mode / force non digitally signed drivers etc.....

if you then hit esc to cancel, it forces the multiboot options screen and gives you the windows memory diagnostic option....

---

i got back into windows...and was in media center watching 'the sarah silverman show'... when i got one error... followed by a freeze...then a BSOD..then a reboot.

I put in my vista disc & went to the repair section. ran the mem diagnostic.
in the "status" section.. it said that the hardware had errors.
when i booted into windows, i didnt get any more feedback. then BSOD.. crash again.

im confused. how could everything run perfect, then all of the sudden, i try to go from 3.0 to 3.4 & now the ram is dead? or is it?

any other way to test the ram?
ugh... im soo frustrated.

Run memtest on each stick and find out which one is bad. I had similar issues yesterday so I ran memtest and one of my sticks were bad. Replaced them today with another set (great support by Team Group btw) and all is good again. :)

Quote:

---

Originally Posted by theonlybabyface

Run memtest on each stick and find out which one is bad. I had similar issues yesterday so I ran memtest and one of my sticks were bad. Replaced them today with another set (great support by Team Group btw) and all is good again. :)

---

ok.. im downloading it now. anything specific for instructions?

what do you think caused this?

It's possible you do have a bad stick but your problems started with a config change so unless you accidentally pumped 4 volts to your memory, they're probably OK.

Reset your BIOS settings to the defaults and re-flash the BIOS. It's possible a setting got screwed up behind the scenes.

You're overclocking...always a risk involved and stuff breaks sometimes. :D

Quote:

---

Originally Posted by argylesocks

ok.. im downloading it now. anything specific for instructions?

---

Just burn iso to a cd, set boot from removable to enabled in the bios and run test 5...if there's a bad stick, you'll find it.

just ran memtest 86+. not quite sure how to read all the info. but im gonna go out on a limb & say that the 13000 errors it found is not a good thing. :(

odd. im 100% possitive, i NEVER had this ram above 2.08v. the specs are 1.9-2.1v.

gtj - i reset all my settings in the bios, and tried to reflash 2634 but it wouldnt work, as it is already on there. at this point, is it worth trying to flash to 2333 or something?


could this be a windows issue?? now various programs are unable to load. but would that cause an orthos crash?
---
update: flashed to 2507. checked all my bios settings. everything is stock on the cpu. on the memory, i put it to 4,4,4,15 @ 2.08 (per ocz). i also changed ref=266 and mem=533
rebooted in safemode. tried orthos. lasted 1 minute.

i went to the ocz forum. people were questioning thier ram. showing screenshots of memtest with tons of errors, and the ocz guys were still suggesting changing voltage, and such. huh? if there are errors, the ram is junk, right?

should eist be disabled and how if i need to? will it give a perfomance boost at all to disable eist?

All this talk about the board killing ram just drove me to drop the voltage back down on my OCZ Plats from 2.04. Passed a couple hours of Memtest at 4-4-4-15 and 1.92v. Don't want to run the risk. Might even drop it another notch.

Quote:

---

Originally Posted by purplehaze

should eist be disabled and how if i need to? will it give a perfomance boost at all to disable eist?

---

Both C1E and EIST are components of Intel's SpeedStep tech designed to drop the CPU multiplier and CPU core voltage (respectively, I think) to conserve power and reduce heat/need for cooling and ultimately fan noise.

You disable them for testing so that when a test like Orthos is running you don't get a drop in power that makes your system unstable. At least that seems to be the theory.

Pretty sure that once you've established a stable overclock you can enable C1E at least (and probably EIST) and still function fine. There seems to be a bug in the bios that gtj has reported to intel where disabling C1E doesn't actually disable it. I've seen my own multiplier drop down at idle in CPUid, and I've also noticed the Vcore drop at idle in both Everest and Speedfan ... from my E6600's stock (with Vdroop) of 1.24v to 1.16 in Vista 64. Looking at it right now actually.

I have a feeling that both C1E and EIST are not being disabled properly.

Regardless, you need to disable them for testing purposes. The most important thing to disable on the new Bad Axe 2 bios is the Watchdog function which will overreact to overclocks and refuse to boot if it's enabled. You can find Watchdog by switching a jumper on the board. It's located in the manual and further back in this thread.

so i gave it another shot at 1:30am. opened orthos.. again get the error message:
Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4

shut down. go to bed for 4hrs of sleep.
as im making my breakfast/coffee, i reboot. run orthos again. this morning, it ran for 31mins fine without problem. i had to shut down to go to work. a couple of programs would not open properly.... maybe another restore point?

im confused. can memory errors correct themselves? maybe my memory was tired & needed some sleep.
http://farm1.static.flickr.com/115/4...336808c0_m.jpg

argyle: Heat, if you turn it off for a while and it runs fine when you get back... you might be getting some type of heat saturation issue,

everyone: new bios 2663
New Fixes/Features:
 Updated processor support.
 Fixed a problem where Setup F9 key caused unexpected page movement
at submenu option.
 Fixed a missing highlight cursor problem in Setup after pressing
F10 and no response.
 Fixed Setup cosmetic issue where garbage characters are printed
outside of string input box.
 Fixed F10 boot menu showing two lines of duplicated title.
 Added AMT1.0 support for certain SKUs.
 Fix POST hang when SATA Controller is set to disable and IDE/ATA
Mode is set to legacy mode.
 Fixed issue where certain host adapter cards not functional.
 Added support for loading the Microsoft Windows Vista* native TPM
1.2 driver.
 Fixed yellow bang in Windows XP* Device Manager for the TPM 1.2
device.

Quote:

---

Originally Posted by stmckin

argyle: Heat, if you turn it off for a while and it runs fine when you get back... you might be getting some type of heat saturation issue,

---

how do i check that? everything was fine up until i messed with it last night. voltages, even at stock settings - i still got crashes.
my core temps were really low.

dunno. i wish there was a giant 'undo' button i could press right now.
;(

gtj - you say power completely down after i change the bios voltage settings... after i F10 (and save), it auto reboots & wants to start windows... is it ok to just hit the power switch during post??

additionally - is there anything wrong with changing any of my settings while in 'maintenance mode' ?? or should i just change watchdog, then reboot in normal mode?

Quote:

---

Originally Posted by argylesocks

how do i check that? everything was fine up until i messed with it last night. voltages, even at stock settings - i still got crashes.
my core temps were really low.

dunno. i wish there was a giant 'undo' button i could press right now.
;(

gtj - you say power completely down after i change the bios voltage settings... after i F10 (and save), it auto reboots & wants to start windows... is it ok to just hit the power switch during post??

additionally - is there anything wrong with changing any of my settings while in 'maintenance mode' ?? or should i just change watchdog, then reboot in normal mode?

---

Yeah, just hit the PSU power switch and no problems changing things in maintenance mode. As someone pointed out earlier, when you do go into maintenance mode, your reference and memory frequencies will get reset so you have to change them back to something sane.

To check the memory, run memtest with only 1 stick at a time. It could also be a heat issue as stmckin pointed out. It's definitely not Windows.

Quote:

---

Originally Posted by gtj

Yeah, just hit the PSU power switch and no problems changing things in maintenance mode. As someone pointed out earlier, when you do go into maintenance mode, your reference and memory frequencies will get reset so you have to change them back to something sane.

---

ok. thanks. yes, i did go back into regular mode & change them to 266/533.

Quote:

---

Originally Posted by gtj

To check the memory, run memtest with only 1 stick at a time. It could also be a heat issue as stmckin pointed out. It's definitely not Windows.

---

i will try that tonight. although, why would there be a heat issue now? im running things the same as before.. and actually back at stock.
im baffled.

They have NOT fixed the C1E-Bug? :(

Could someone please check on that?

I cant, im using a Pentium 4 630 right now, traded my E6300 for an E6600ES.