[d94]

for 65nm cpu's
had a random idea to start this thread LOL
and if it gets popular i will make a list in this post :P
to get in the "club" you just need to post a cpu-z verification link..and a screeni if you feel like it

mine:

comments/suggestions?

members:

d94 920 4515 stable pcmark 4806 max
ross 661 5318 stable pcmark 5678 max
batboy 940 4617 stable
markodude 955xe 5031 stable pcmark

[Ross]

Aww, now I am wishing I at least validated one of my 930's 4.8GHz screens

Oh well, here's the Cedar Mill 661 "so far"...

5678 (1.58Vc)



PCMark04 (less voltage than above 1.55Vc...working on higher here)

[batboy]

Any rules as to stability? Heck, anyone can post a super high suicide run CPU-Z screenshot. I say you have to be at least PCmark2004 stable. So, that means I'm at 4.6 gig with the "green" rig in my signature.

http://valid.x86-secret.com/show_oc?id=64638

My PCmark2004 screenshots and suicide run CPU-Z screenshots are all in my thread if people really wanna see 'em.

[dylskee]

Reserved........... My Cedarmill will be here in two more days!!!!!!

[d94]

Originally Posted by batboy

Any rules as to stability? Heck, anyone can post a super high suicide run CPU-Z screenshot. I say you have to be at least PCmark2004 stable. So, that means I'm at 4.6 gig with the "green" rig in my signature.

http://valid.x86-secret.com/show_oc?id=64638

please post voltage as well

[Ross]

I agree that PCMark04 WITH A SCORE screenie (link when Futuremark catches up) should be included also to differentiate useable from not. That will also prove or disprove my theory that CM cores are stable very close to their max

[d94]

Originally Posted by Ross

I agree that PCMark04 WITH A SCORE screenie (link when Futuremark catches up) should be included also to differentiate useable from not. That will also prove or disprove my theory that CM cores are stable very close to their max

that sounds good to me
should each member of the club post say a stable oc and a max oc?
for example:
d94 4515 stable 4800 max

this would mean the max oc would need a validation link..or should we just say a cpu-z screeni?

[Ross]

Definitely both. I like seeing screens too, but a link should be included for verifications for CPU-Z and PCMark04.

[xTrEmEoVrClOcKr]

Wondering if the Preslers will hit 5ghz on air / water .. could be a possibility if chips like D94s get better w/ steppings and revisions.

[pscout]

Assuming I get to putting water on my 920 this weekend, I might post here again.

So far I could only get to 4.346 pcmark04'ed at 1.4875 vc on air

Then I will have to learn how to get validated results too.

[batboy]

Just go to the "about" tab of CPU-Z and click on "validation" button, then follow the instructions.

I tend to think if the air cooled folks can hit a max screenie, they should be allowed to join the club even if they aren't stable.

[Ross]

I personally would like to still see both PCMark04 for air here just to make it complete. I say if they suicide @ 4.5GHz+, post that CPU-Z and a PCMark04 anyway even if it's only 4GHz...

[pscout]

Originally Posted by batboy

Just go to the "about" tab of CPU-Z and click on "validation" button, then follow the instructions.

I tend to think if the air cooled folks can hit a max screenie, they should be allowed to join the club even if they aren't stable.

Ok so now i know how to do validating ... and that it took vc=1.5625 to get the cpu-z screenie at 4.514 ghz, but sadly no pcmarko4 ... reboots during test 5.

So I will just have to wait for water.

[Ross]

PCMark04 should be either a significantly less clock or significantly greater Vc to finish with a score compared to suicice! It's tough. I am surprised it actually made it to test 5. Try 4.2-4.3GHz and see if it will complete there with a score.

Hint: if during the concurrent tests (first 3 or so), if one of the bars stops, you will not get a score, even if the other continues and it completes the bench. If that happens, you might as well save some time by aborting and make some adjustments

[pscout]

Originally Posted by Ross

PCMark04 should be either a significantly less clock or significantly greater Vc to finish with a score compared to suicice! It's tough. I am surprised it actually made it to test 5. Try 4.2-4.3GHz and see if it will complete there with a score.

Hint: if during the concurrent tests (first 3 or so), if one of the bars stops, you will not get a score, even if the other continues and it completes the bench. If that happens, you might as well save some time by aborting and make some adjustments

I completed a pcmark04 yesterday at 4.346 at 1.4875 vc. It has been dual folding since last night at /edit 290 not 300 ... completed 2 QMD's successfully already ... so far it has the highest folding ppd i am aware of for 1 rig (915 ppd).

Since my bios temps at 1.5625 were 58, i decided i will just wait till i put water on it so i have put it back to work on the folding farm.

Although once booted temps were ok at 42 idle ... peaked at 64 in some of the early pcmark04 tests.

[proth]

Pardon me while I post this as a place holder..

Originally Posted by Ross

Aww, now I am wishing I at least validated one of my 930's 4.8GHz screens

Oh well, here's the Cedar Mill 661 "so far"...

5678 (1.58Vc)

"so far"??? Hmm. I see 1.58v up there. How long did you run it that high?

[maxxpehling]

ross's max score says 5.8 instead of 5.8, will get mine in a week and a half or so! I'm going to play with stock cooling for a few days and then put my zalman 7000 on it should be interesting

[Ross]

ross 661 5318 stable pcmark 4678 max

Yeah, stable at 5318, but only maxed at 4678, LOL. Wanna fix that first digit in the max clock?

[d94]

Originally Posted by Ross

Yeah, stable at 5318, but only maxed at 4678, LOL. Wanna fix that first digit in the max clock?

done :P

[hUMANbEATbOX]

ahh i was way ahead of you guys...i shot d94 a pm about that yesterday

nice clocks guys, can't wait till this club grows annd grows and grows....

with any luck i'll be in before my birthday (apr).

[markodude]

Currently at 4.0ghz prime stable 1.30v actual, had her benching at 4.2 but this board craps out above 240 with all the dual cores I have tried.

Will be joining this club when my new board arrives.....hopefully

[Ross]

What I will be curious about is what improvement, if any, the -Es are going to offer now that they are available. I am fighting ordering one because I really don't think it's big enough of a upgrade at this point.

[wingman99]

Originally Posted by d94

for 65nm cpu's
had a random idea to start this thread LOL
and if it gets popular i will make a list in this post :P
to get in the "club" you just need to post a cpu-z verification link..and a screeni if you feel like it

mine:

comments/suggestions?

members:

d94 920 4515 stable pcmark 4806 max
ross 661 5318 stable pcmark 5678 max
batboy 940 4617 stable

I have been reading about thermal hot spots on the 65nm have you guys been using ThrottleWatch to make sure your scores are at the MAX. You might be able to score highenough to beat AMD's top score in PCMark04 of (10089) and take the crown away from AMD

http://service.futuremark.com/orb/projectsearch.jsp

Maybe use VapoChill, dry ice,CO_2, or insane Liquid Nitrogen

[Ross]

I have not used TW.

A Presler probably stands every chance of topping 10K in 04 if you can get one 04 stable @ say >4.6GHz and have a good setup. My 930 did 9200 @ 4.3GHz with crappy components. With a great vid card, SATA RAID and a streamlined environment, I think 10K+ will be had around 4.6-4.8GHz on a Presler no problem and that might not even take anything more substantial than a single stage or even good water with cool ambients. Dice, cascade or LN2 and I dare say it would probably be "easy".

[Npvk_x]

"booked"
by the end of the week my new mb shoudl arrive and i can find out just how much my Targan 580w killed... hopefully just the mb

will post my 5ghz validation if my hdds didnt also die... trying not to get my hopes upp..

[dylskee]

I just noticed the title said 5GHz for the CM, i don't think i'm going to need my reserved spot way up there^^^^. I just saw the 4.5GHz and reserved my spot. Oooops, i'm on air so i'm not expecting 5GHz, but i can't wait to try!

My chip has sitting at the facility 8 miles away since 9:00 AM this morning and they didn't deliver it today! Oh well, will be here tomorrow but i won't be home till 6:00 PM!

[Ross]

d94, I need you to update my stable from 5318 ->5407....PCMark04 is below.

This score is nuts for a single core with a crappy vid card/no RAID/no ram drive or I-RAM. Took 1.60Vc to get the score and I am still trying to do it DDR1000 4-2-3-4 for >8200.

If this were a Presler, it'd be like 14000



The 661 is also kicking the crapola out of my 670...benching SPi 1M about where the 670 suicided. 5570 with a whopping DDR1030 4-2-3-4. 661 has 150MHz higher clock, 100MHz faster DDR and 80FSB LESS than the 670 and it beat that best score by .4s.

More than anything, it's a testament to their drop in TDP and stability near max clocks. The 670 on phase struggled to complete a SPi 1M @ 5.4GHz before heat became an issue on the SS. I think the CM could easily do 5.6GHz with the extra Vc it will take to do it on the same SS.

[proth]

Unbelieveable performance Ross. Seems like you have that thing tuned just right.

Hmmm.... Yesterday you ran 1.58, Today 1.6, wonder what tomorrow will bring.


Be careful with that Vcore. Think how you will feel if it pops

EDIT:Can you please post your load temps @ 1.6
thx

[Ross]

LOL...yeah, who knows. I wouldn't like it if it pops and I am nearing my limit...I'm really thinking about buying either a 631 or 920 for the sole purpose of thrashing within an inch of its life...this 661 is too nice for that We know the 920s look a lot better than the others, but I'd like to see 631 in action first. Since there's no 621, it should be OK, but like I've said before, I have no luck with any Intel proc that has a "3" in it...like the 630 and the 930.

I have no idea what my temps are under load...definintely toward the positive side, but since it only goes as low -16C on idle, I am not surprised. I don't believe the temps is reads on phase at all, but they can't be too bad or it would never finish PCMark04 there from lack of capacity...the 670 was limited to 5GHz on all 3D/PCMarks for that very reason. They only thing it could do >5GHz really was SPi 1M because it's so short...

[speed bump]

very nice ross don't forget to post your time in the superpi thread.

I think I am going to give pressler a shot though, just as soon as I find a GPU and a good price on a 920.