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

[slappynutz]

My E6600 needed at least 1.5250 vcore to be stable at 400 FSB with the same "up 2 notches" on the vmch and vfsb. Actually that's where "ralph" is now so I guess I need to update my sigs.

Thanks. Was hoping I wouldn't have to go too much over 1.4000 vCore to hit that magic 400 FSB mark. What sorts of temps do you register at 1.5250 in your P180? Just curious ... we have almost the exact same config. I don't really want my idle to get over 40c or my load to go over 60c.

I should probably just turn the temp monitor off entirely. It just makes me obsessed with lowering already low temps that are well within spec.

[Zucker2k]

I guess this answers the question. Revisions are not only bios related.

New PCN, board revision 506.
Description of Change to the Customer:
1. Updating the BIOS to revision 2507.
2. Adding the option to use both Maxim parts and the existing Pericom
parts in PCIe x8/x8 switching circuit.
3. Adding an enhanced revision of the ADI voltage regulator controller.
Additional items for retail desktop boards:
a. China RoHS documentation enhancements.
b. Updated printed product guide.
c. Updated the driver CD.

[vorpoxtrm]

Thanks. Was hoping I wouldn't have to go too much over 1.4000 vCore to hit that magic 400 FSB mark. What sorts of temps do you register at 1.5250 in your P180? Just curious ... we have almost the exact same config. I don't really want my idle to get over 40c or my load to go over 60c.

I should probably just turn the temp monitor off entirely. It just makes me obsessed with lowering already low temps that are well within spec.

I don't know if this will help, but I also use a P180 case with a Tuniq. Using a E6600 636A CPU, @400 FSB I need 1.4375V Vcore to get 9+ hours Orthos Blend (Priority 4) stable. Coretemp @ idle is 35C, @ orthos load 55C, TAT 100% 63C with 22C ambient. The highest I've been able to get 9+ hours orthos stable is 412 FSB with 1.475V vcore. After that no matter what vcore was set, there were stability problems for my setup.

[slappynutz]

I don't know if this will help, but I also use a P180 case with a Tuniq. Using a E6600 636A CPU, @400 FSB I need 1.4375V Vcore to get 9+ hours Orthos Blend (Priority 4) stable. Coretemp @ idle is 35C, @ orthos load 55C, TAT 100% 63C with 22C ambient. The highest I've been able to get 9+ hours orthos stable is 412 FSB with 1.475V vcore. After that no matter what vcore was set, there were stability problems for my setup.

Cool, thanks. That's about what I was seeing last night.

I'd love to get my E6600 to 400FSB if only to get my memory clocked at its rated speed. Starting to wonder how much of a performance jump I'll really be seeing between 3.24ghz and 3.6ghz. Wish there was an overclock database with different benchmark specs out there somewhere. Haven't been able to find one.

The CPU box states my max voltage as 1.35 and the CPU vid reports at 1.214 in CPUid (or somewhere thereabouts). Guess it's just newb angst keeping me from going over 1.35 actual voltage in Windows (accounting for the .05/.04 vdroop).

[Blacklash]

If you're having problems with a FSB beyond 411 on a 9x multi, at some point you may want to shift to a 333 ref and set a freq of 667. That will put you @ 1:1 on a 1333 strap.

I can boot and briefly bench @ up to 3.9 (434 x 9) and I have to shift to a 1333 strap to be stable. If I stay on a 266 ref I am limited to 3.8. This chip needs 1.65v to move beyond 3.9 and that's not available on the BX2.

[gtj]

Cool, thanks. That's about what I was seeing last night.

I'd love to get my E6600 to 400FSB if only to get my memory clocked at its rated speed. Starting to wonder how much of a performance jump I'll really be seeing between 3.24ghz and 3.6ghz. Wish there was an overclock database with different benchmark specs out there somewhere. Haven't been able to find one.

The CPU box states my max voltage as 1.35 and the CPU vid reports at 1.214 in CPUid (or somewhere thereabouts). Guess it's just newb angst keeping me from going over 1.35 actual voltage in Windows (accounting for the .05/.04 vdroop).

Don't worry about the voltage.

If you really want to see what you're rig will do, set as follows...
vCore 1.5500
vMCH 1.6
vFSB 1.4
Ref/Mem 333/667

Watch your temps of course. If this gets you stable at 400 and your temps are acceptable, try moving the FSB up. If you're stable but your temps are too high for your liking, move the voltage down until you're not stable anymore at 400. Etc, Etc.

[gtj]

gtj, did you have a chance to look at my above post. Any concerns over the temps I am seeing at 350X9 on my E6600?

Is it even worth trying for 400X9 at this point?

Yeah, they are a little high but not too bad. As Argylesocks says, try remounting the Zalman. Which way is it oriented?

[minsc_tdp]

I'm now at 380 FSB stable for 12 hours blend at 1.45V CPU. Going for 400 tonight. I was misled by the unstable ATITool crashes which I attributed to an unstable system and had my voltage way up at 1.5 but it wasn't necessary. At some point also, my memory decided to change itself from 533 to 667 which might have messed me up, watch for that too.

EDIT: make that 390, still 1.45V

[jmpage2]

Yeah, they are a little high but not too bad. As Argylesocks says, try remounting the Zalman. Which way is it oriented?

I've just re-installed the Zalman 9700 three times and tried very light coatings of both Arcit Silver Ceramique as well as the Zalman Thermal Grease that ships with the 9700.

Nonetheless my temps seem unusually high.

Zalman bracket for the cpu socket is definitely installed correctly and all four screws tightened down firmly.

I am using the "horizontal" installation with the Zalman fan pointed at the front of the case so that the airflow will exit out the back of the case and the "s" bracket for the 9700 installed horizontally so that one screw is on the right side of the bracket and the other on the left.

It's a little bit warm in my office, probably around 23C. Nonetheless, after every single installation I have done the CPU shows that it is idling around 40C+ in TAT and is hitting as much as 65C for brief periods during Orthos testing. It just seems unusually high. I have double checked and the Zalman cooler definitely seems to be making good contact with the heat spreader on top of the CPU and I'm not using too much thermal grease.

TAT is reporting right now my idle temperature being 44C and the System Zone 1 and 2 showing 47 and 48 degrees respectively.

I'm wondering if this is related to the relatively poor air flow afforded by the Gigabyte Poseidon case. I know other people with this case have complained about the airflow being slightly sub-standard.

One thing is for certain, I can't contemplate going any higher with temps this high.

[gtj]

TAT is reporting right now my idle temperature being 44C and the System Zone 1 and 2 showing 47 and 48 degrees respectively.

I'm wondering if this is related to the relatively poor air flow afforded by the Gigabyte Poseidon case. I know other people with this case have complained about the airflow being slightly sub-standard.

Bingo!!

[jmpage2]

Bingo!!

Any solutions that don't involve swapping out the case?

[gtj]

Any solutions that don't involve swapping out the case?

Well, I see by the description it's got 2 "silent" fans. Are they adjustable? If not, that'd be the first thing I'd change.

I also see 2 "liquid cooling ducts", hint, hint.

[slappynutz]

Don't worry about the voltage.

If you really want to see what you're rig will do, set as follows...
vCore 1.5500
vMCH 1.6
vFSB 1.4
Ref/Mem 333/667

Watch your temps of course. If this gets you stable at 400 and your temps are acceptable, try moving the FSB up. If you're stable but your temps are too high for your liking, move the voltage down until you're not stable anymore at 400. Etc, Etc.

I was wondering about the strap. You recommend going up to the 1333? I'd heard there might be issues doing that, and noticed you were still on the lower strap on your test system so I left my system at 266/533. I have the exact same ram you do only running at a lower voltage (1.92v). Just wondering.

[jmpage2]

Well, I see by the description it's got 2 "silent" fans. Are they adjustable? If not, that'd be the first thing I'd change.

I also see 2 "liquid cooling ducts", hint, hint.

Well, I just replaced the clear window on the side of the case with the steel mesh one that was also in the box. This alone has dropped the temps 8C which definitely points towards air flow being a culprit.

Now peaking out on Orthos at around 56C under full load and idling in the 30's again.

Not sure if I will go much further though with air cooling on this case, I don't want it to sound like a jet engine sitting on the desk. 3.15ghz is still not shabby in my book.

[SiGfever]

Well I took a posters advice and now have read all 102 pages of this thread. I have read every post and every reply and I have one thing to say...You guys ROCK!

There are so many talented and helpful people in this thread that this is what separates this forum from a lot of others. I am letting my TIM cure some more before I have at it heavy, I have been up to 3.4Ghz but backed down to defaults. I really appreciate the help that you guys give and will definitely be seeking your advice.
Thanks...

[Supertim0r]

gtj, you're really the XBX2 king

I haven't played with my mobos for weeks

[SiGfever]

@Supertim0r,

I notice in your sig it says "Formation in progress...refrigeration technician"

I am a thirty year Master HVACR tech.

[gtj]

I was wondering about the strap. You recommend going up to the 1333? I'd heard there might be issues doing that, and noticed you were still on the lower strap on your test system so I left my system at 266/533. I have the exact same ram you do only running at a lower voltage (1.92v). Just wondering.

Yeah, I didn't finish my thoughts there, sorry. Those settings are an almost fool-proof starting point not the optimal point. If you're stable there, then it's up to you to decide which things are more important for the types of work you do and what kind of compromises you're willing to make. Voltage vs Temps vs CPU Speed vs Latency vs Stability vs Memory Speed, etc.

Keep good records of what you've tried and experiment!

[gtj]

gtj, you're really the XBX2 king

I haven't played with my mobos for weeks

It's amazing what being out of work allows one to do.
We'll see what happens in a few weeks when I go back.

[argylesocks]

I am using the "horizontal" installation with the Zalman fan pointed at the front of the case so that the airflow will exit out the back of the case and the "s" bracket for the 9700 installed horizontally so that one screw is on the right side of the bracket and the other on the left.

im not sure if it makes a difference (doubtful), but i believe that "S" shaped bracket is actually supposed to be mounted vertically. I did it horizontally at first, and then realized the instructions say if to mount it the other. Again - not sure that it makes a difference... but worth re-reading the instructions. The pictures are horrible... but thats how i understood it.

http://www.arcticsilver.com/pdf/appi..._dual_wcap.pdf
check out page 4 for a picture of the Artic Silver


edit: after re-reading the directions (for the twelveth time) i believe it should be horizontal... sorry. im a dumba$$

[gtj]

im not sure if it makes a difference (doubtful), but i believe that "S" shaped bracket is actually supposed to be mounted vertiacally. I did it horizontally at first, and then realized the instructions say if to mount it the other. Again - not sure that it makes a difference... but worth re-reading the instructions. The pictures are horrible... but thats how i understood it.

http://www.arcticsilver.com/pdf/appi..._dual_wcap.pdf
check out page 4 for a picture of the Artic Silver

Good catch. With the bracket horizontal, the weight of the radiator and fan could be pulling the base off of the cpu slightly.

[argylesocks]

This doesnt happen all the time, but at least 50% of the time. When my computer comes OUT of sleep mode, it often reboots itself when i start up media center... or if i was in MC, when it went to sleep, it will start to boot, then reboot again...

I know the beta version of the Nvidia drivers solves "GeForce 8800 GTX/GTS: Display does not resume from S3 or S4
suspend mode", but based on other's opinions, i have not installed these drivers.

I have have installed the Viiv 1.6 software, as well as the Intel Quick Resume software (and turned on in the bios).

running vista x86. thoughts?

[slappynutz]

Don't worry about the voltage.

If you really want to see what you're rig will do, set as follows...
vCore 1.5500
vMCH 1.6
vFSB 1.4
Ref/Mem 333/667

Watch your temps of course. If this gets you stable at 400 and your temps are acceptable, try moving the FSB up. If you're stable but your temps are too high for your liking, move the voltage down until you're not stable anymore at 400. Etc, Etc.

Did as you suggested, bumped the FSB to 380 and started testing. My load and idle temps jumped about 10c, with the max under load being 60/62c. Is that normal? I backed back down to 3.24 as I'm starting to wonder if a handful of extra unused fps are worth the increase in temps and fan noise.

[jmpage2]

Good catch. With the bracket horizontal, the weight of the radiator and fan could be pulling the base off of the cpu slightly.

That's a good point. Actually the documentation for the 9700 on socket 775 specifically shows the "s" bracket installed horizontally with the two lugs to the right and left of the fan. I suppose I could try changing the bracket around and see if that has any effect on temps.

[ChaosMinionX]

Anyone have a problem where the bios takes awhile to post? All of a sudden its just taking awhile to post, it sits at the bar from load and waits awhile to show the Intel screen. Any ideas?