Asus P6T Deluxe Discussion Thread

[MaleX]

Has anyone tried the new 0704 bios for the P6t Deluxe V2?

Yeah I tried 0704 and it was horrible!

Windows would hang at the animated logo at startup when it had the exact same settings as 0610. Also at a certain point it could not save any new settings anymore forcing me to turn it off/on again.

Flashed it back to 0610.

[unclewebb]

I had a wierd bug with realtemp this run.

That's a new one. Prime stable but not RealTemp stable.

I just sent you a link to the latest version of RealTemp. Hopefully it works OK for you. There have been lots of changes and hopefully improvements since version 3.30.

[nkresho]

^ priming with it right now. Thanks. Looks fine to me so far.

[heydo6]

I've since removed the nb fan. I pulled it and found it doesn't get warm enough (to the touch) to need it on there. At least, at my speeds so far. And it was buzzy sounding (i hate tiny fans).

Cheers y'all![/QUOTE]

water block looks to be on up side down ? inlet in the middle and outlet up top so theres no air pocket ....

[infected_]

Did you have to remove the battery or clear the CMOS after flashing it to 0006?

No. Bios defaults are loaded automatically, because of the commands used in the afudos.

What is the easiest way to roll back to a prior bios?

I am not happy with 0704 (some strange system slowdowns since I loaded it).

use this afudos, place in a bootable usb stick with the bios, and use: afudos /iROMNAME.ROM /pbnc /n

[infected_]

Anyone know if its better to buy a recent manufacturing week of w3520's or older weeks?

[nkresho]

water block looks to be on up side down ? inlet in the middle and outlet up top so theres no air pocket ....

Didn't even realize that. Thanks for the find. I'll have to flip it over.

[stutheterrorist]

Hey unclewebb, sorry for time taken to get back about this...

stutheterrorist and other users. I have been talking with Asus and they are becoming skeptical about some of the reported turbo throttling. They don't fully trust the i7 Turbo application I wrote and would be willing to believe more if CPU-Z is also in your screen shots showing this same problem. In my opinion, recent versions of CPU-Z that rounds the multiplier off to the nearest whole number is not an adequate tool for the job but my opinion doesn't count.

Can you post some screen shots with the latest 0001 special bios that includes i7 Turbo and CPU-Z during some extreme turbo throttling so both programs clearly show this issue that would be great. Their engineers seem to think that if CPU-Z doesn't show it then it must not be happening.

i7 Turbo is not something I cooked up off the top of my head. It follows the method outlined in the Intel Turbo White Paper that came out last year. If Asus doesn't trust my program then they must have at least one programmer on staff that can write a similar utility in the matter of one day. It's not an overly complex programming exercise.

http://download.intel.com/design/pro...ots/320354.pdf

Intel's paper makes it clear in section 2.1.

2.1 Problem with Today’s Software Reporting Frequency

"....., software that uses the Pstate structures, directly or indirectly, will not properly display Intel® Turbo Boost technology frequencies."

I have followed Intel's recommended method as outlined in this paper while most other programs do not.

For sure, more than happy to and then they'll see ha-ha-ha !! Will do best to get round to it over the weekend if possible but will definately get it done soon as..

I flashed back to the V2 0001 bios but this time there was no turbo throttle !!

Also, tried going back to normal Asus bios 0610 and redoing what I did first time but again, no throttle with the V2 0001 bios on first flash to it.

Am still 100% sure it did throttle first time as that's why I flashed to the V1 0006 bios and posted my comments... Tad embarrasing now ha-ha-ha !!

Hmmmm... Perhaps a iffy flash first time of trying it ?? Don't know but the V2 0001 bios does work... (Also got the turbo logs if wanted.)

[unclewebb]

stutheterrorist: Maybe you are right on the edge with your present settings. Prime95 Small FFTs or LinX which uses more load might be enough to push you over that edge. Your core temperature might be a few degrees cooler with your recent tests as well. I don't know either. If 0001 works for users, great. If it doesn't work, there's always 0006 for the v1 waiting to be cross flashed on to a v2 board. 0006 is a proven performer.

[stutheterrorist]

stutheterrorist: Maybe you are right on the edge with your present settings. Prime95 Small FFTs or LinX which uses more load might be enough to push you over that edge. Your core temperature might be a few degrees cooler with your recent tests as well. I don't know either. If 0001 works for users, great. If it doesn't work, there's always 0006 for the v1 waiting to be cross flashed on to a v2 board. 0006 is a proven performer.

I got temps to 86 84 82 81C using LinX with max memory and 15 passes, volts were set to 1.3625 in the bios. Under Prime blend they got to 84C. Didn't use the Small FFT's. Will give it a try.



I'm assuming the slight multi drop was due to pic being taken shortly after the run had finished, about 30 seconds to a minute after.

I'm gonna keep using it and see how it plays out.. If not will go back to the v1 0006 for sure.

[infected_]

unclewebb, i read stutheterrorist post, and decided to try the 0001 bios again, but using afudos and the crossflash commands...



shouldn't already be throttling at that voltage and temps?

Edit: i think its not really throttling!

[unclewebb]

S1N3 posted a screen shot and sent me a log file that showed turbo throttling when bios 0001 was first released.
http://www.xtremesystems.org/forums/...postcount=1401

The log file didn't show a one time random reporting error or anything like that. It was very clear.

I have no idea why 0001 didn't seem to work but now it does. I don't know if this is some sort of intermittent problem or what the deal is.

If 0001 works for you then use it. I've never seen a screen shot of 0006 throttling.

[infected_]

im not saying your program reported it wrong or anything like that. i posted also a screen of the turbo throttling when the bios 0001 was released, a few pages back.. could it have to be the way we flash it?

[unclewebb]

I don't have any reasonable explanation for turbo throttling being there one day and not the next.

I don't think the flashing method would have anything to do with it but

I've seen some motherboards do unpredictable things after coming out of stand by mode but I don't know if that played a factor in this either.

At the moment, it goes in the sh_t happens category. It might be difficult to figure out the exact reason. Maybe a few more users will give 0001 a second try now and report their results. Your results look great.

[S1N3]

Yeah I tried 0704 and it was horrible!

Windows would hang at the animated logo at startup when it had the exact same settings as 0610. Also at a certain point it could not save any new settings anymore forcing me to turn it off/on again.

Flashed it back to 0610.

It would be interesting to know if you have a SSD, seems a few folks including myself have had issues with 0704 and SSD's.

[nkresho]

I was getting very obvious (in i7turbo as well as in cpu-z) throttling at 21x205 4300mhz.
I just afudos flashed to 0006 as per uncleweb's posts.
Running lin-X at my 4200 speed to make sure it's still stable.
I'll post results if I get it going without throttling at 4300.

BTW, this thread is awesome.

[HDCHOPPER]

It would be interesting to know if you have a SSD, seems a few folks including myself have had issues with 0704 and SSD's.

I would like to know this also big time as I use SSD's

[stutheterrorist]

BTW, this thread is awesome.

+1...

[nkresho]

Well it's looking more like a limitation of my chip than anything else.

I was able to get into windows at 4300 and I was able to run lin-X at 1.38750 cpu and 1.350 QPI.

I got into windows at 4400 but the increase in voltage made me a little nervous. I was still getting linx errors at 1.41250v cpu and 1.38750 qpi. I had memory relaxed to 1260 too.

I think i'm going to back her down to 4200 and try to be happy with it. If I get the itch, and I will, I'm going to have to find another chip, cause this one isn't cutting it for me. I was able to do 4200 at 1.33125 cpu and qpi.

In all honesty, it doesn't seem any faster than my q9650 was at this speed. In 3dmark vantage, at the same settings, I'm getting roughly the same score. Actually slightly less.

w3520 P17443 http://service.futuremark.com/compare?3dmv=1664503

Q9650 P17539 http://service.futuremark.com/compare?3dmv=1303529

And it's reading my w3520 at 4000mhz when it was actually at 4200?

Same oc for the gtx: 765-1742-1400 in both cases

[Electric Wizard]

So after a week of trying to get 4.4 stable I realize that one of the contributing factors may have been I Was paying so much attention to the CPU I didnt seem to notice one of my ram slots was dead, ugh.

[nkresho]

^ That really sucks. ^ I feel your pain.

If you're planning on rma'ing it, please post back here as to the timeframe they took to rectify the issue. I'm curious and i'm sure others are too.

[unclewebb]

And it's reading my w3520 at 4000mhz when it was actually at 4200?

Some of the FutureMark apps ignore turbo boost so it thinks you're only using the 20X default multiplier.

[nkresho]

Ok, I was wondering. That makes perfect sense.

Even though. Shouldn't my 3dm vantage score be higher? To compare those runs. I wqas running vista x32 on the older test and I'm on 7 x64 now, and I was running 2gb ddr2 whereas I have 6gb ddr3 now.

I have to wonder now, if something is misconfigured?

[MaleX]

It would be interesting to know if you have a SSD, seems a few folks including myself have had issues with 0704 and SSD's.

I'm not using any SSD disks in my system. After flashing back to 0610 everything immediately worked again.

[Raptor-X]

Well it's looking more like a limitation of my chip than anything else.

I was able to get into windows at 4300 and I was able to run lin-X at 1.38750 cpu and 1.350 QPI.

I got into windows at 4400 but the increase in voltage made me a little nervous. I was still getting linx errors at 1.41250v cpu and 1.38750 qpi. I had memory relaxed to 1260 too.

I think i'm going to back her down to 4200 and try to be happy with it. If I get the itch, and I will, I'm going to have to find another chip, cause this one isn't cutting it for me. I was able to do 4200 at 1.33125 cpu and qpi.

In all honesty, it doesn't seem any faster than my q9650 was at this speed. In 3dmark vantage, at the same settings, I'm getting roughly the same score. Actually slightly less.
w3520 P17443 http://service.futuremark.com/compare?3dmv=1664503

Q9650 P17539 http://service.futuremark.com/compare?3dmv=1303529

And it's reading my w3520 at 4000mhz when it was actually at 4200?

Same oc for the gtx: 765-1742-1400 in both cases

I am noting your Vantage CPU (W3520) score is 60105 compared to your Vantage CPU (Q9650) score of 51464. This implies that the i7 acrh is performing 16.7% better than the core 2 arch on a clock per clock basis within Vantage. I do not believe that your "problem" is on the CPU side. +16% is in the expected range.

I believe that your problem most likely is in the OS/ Graphic driver domain..
Regards,
Raptor

Edit:
You should also consider the possibility that Vantage may be GPU limited by your single 285. Adding additional CPU resources to this equation may not help...