E6600 WR modPI1M 9.563s@5304MHz

Printable View

Show 100 post(s) from this thread on one page

03-25-2007, 04:33 AM
fredyama

Quote:

Originally Posted by before

Disabling a core can or can't help... I mean, if you disable a core, the load is focused on only 1 core, but with 2 cores enables, the whole load is shared between core #0 and core #1... Then, it depends on the quality of both cores: if the 2 cores are of equivalent quality, disabling one of both isn't the good way... Spi is a single-threaded app, and with 2 cores enable, the shared load can help getting the max benchable clocks.

Thank you so much for your through explanation, before-san:toast:
Come to think about, New Beetle may have tried with one core at the time of Yonah benching.
Anyway, I totally agree with your advice on this issue;)
03-25-2007, 07:22 AM
alexio

Quote:

Originally Posted by before

Disabling a core can or can't help... I mean, if you disable a core, the load is focused on only 1 core, but with 2 cores enables, the whole load is shared between core #0 and core #1... Then, it depends on the quality of both cores: if the 2 cores are of equivalent quality, disabling one of both isn't the good way... Spi is a single-threaded app, and with 2 cores enable, the shared load can help getting the max benchable clocks.

Very true, but I think it most cases it helps a little to use only one core, but that's just a prediction based on few facts.

I really like this program called "SMP Seesaw Pro". It let's you isolate a process to one core and set the affinity of program-groups with a single click. It gives me a little better time in Superpi (almost unmeasurable) with a P4 HT CPU, but it could be more useful for dual-core CPU's benching on the edge of stability. It's certainly worth a try ;)
03-25-2007, 07:47 AM
Misko

@before

@PRTeamJapan

When I disable one core via win or bios, how CPU disable it , always same core, or he sometimes disable one and other time other one?
03-25-2007, 07:52 AM
SIOUX

congrats...amazing results with a E6600 :toast:
03-25-2007, 01:08 PM
G H Z

Great benching from NewBeetle, havn't seen you guys for awhile :)
03-25-2007, 04:55 PM
Cpt Twitchy

Great to see you posting again Fredyama-San.

Is NewBeetle's pot solid or brazed?

Regards,
Tim
03-25-2007, 05:13 PM
Stapler

Pretty sure it's solid, the Japanese were the first to make solid pots IIRC.
03-25-2007, 05:58 PM
erwinz

very nice clocks.. :) congrtaz.. :D
03-25-2007, 07:30 PM
Revv23

Nice to see some high end benching on non high end hardware!

Very nice new beetle and very nice team japan!
03-26-2007, 05:44 AM
before

Quote:

Originally Posted by PRTeamJapan

Thank you so much for your through explanation, before-san:toast:
Come to think about, New Beetle may have tried with one core at the time of Yonah benching.
Anyway, I totally agree with your advice on this issue;)

You're welcome fredyama-san :toast:
03-27-2007, 12:02 PM
fredyama

From New Beetle

Thank you so much all:toast:

Got a slightly improved 8M time at 2m04.328s with not a better mem timing:(
I thought my 2 x 1GB memory were long gone but somehow came back to life again after more than 1 week rest:D

I may try what results I can get out with 1-core;)

http://homepage2.nifty.com/NewBeetle/8M_124s328.png
03-27-2007, 12:11 PM
Flow

very nice result ... respect.

One question: What clock does this cpu reachs watercooled @ 1,4V ... or how much does it need for 3,6ghz!

Thank you .. just need the information to know how to select the E6600
03-27-2007, 12:27 PM
fredyama

Quote:

Originally Posted by Flow

very nice result ... respect.

One question: What clock does this cpu reachs watercooled @ 1,4V ... or how much does it need for 3,6ghz!

Thank you .. just need the information to know how to select the E6600

Thanks:)
Please refer to Rank #4 Core 36-GP @tanuki for his record-- 1.45Vcore under watercooled to get 3.6GHz.
You may be interested in the CPU Lot No column though not a 100% guaranteed, as always:D
03-27-2007, 12:56 PM
blind_ripper

nice new beetle .
lol had same thing with my G.skill rams one death after a week it woke up again just befor i wanted to do rma :D
03-28-2007, 08:25 AM
fredyama

>Flow: What clock does this cpu reachs watercooled @ 1,4V ... or how much does it need for 3,6ghz!

Additional inf from New Beetle

I have tested 5 x E6600 so far with Commando on aircooling.
Please note that those Vcore were set in Bios not actually measured and room temp were between 18C--20C.
　　　Lot　　　　 1.350V-MAXClock　　1.600V-MAXClock　　　　x6-MAXFSB
　#1(L629A526)　　　　　3618　　　　　　　　3951　　　　　　　　 557
　#2(L629B948)　　　　　3246　　　　　　　　N/A　　　　　　　　 430
　#3(L633A808)　　　　　3690　　　　　　　　4095　　　　　　　　 528
　#4(L633A808-2)　　　3528　　　　　　　　3942　　　　　　　　 500
　#5(L628B127)　　　　　3609　　　　　　　　N/A　　　　　　　　 545

#1 & 5 were only CPU I could OC to over 5G and got U10s of modPI1M under LN2.

In order to OC over 5.3G under LN2, my guess is to find one of E6600 with:
1. capable to go over 3.6G at 1.35Vcore on air,
2. capable to reach over 560 of x6-MaxFSB

It is a kind of luck of draw, i guess:D

Hope you can find a good one:)
03-28-2007, 09:01 AM
jozzy

>PRTeamJapan : :toast: :toast: very nice setup.. may i see the SPi 4M please? :clap:
03-28-2007, 09:25 AM
before

Quote:

Originally Posted by PRTeamJapan

>Flow: What clock does this cpu reachs watercooled @ 1,4V ... or how much does it need for 3,6ghz!

Additional inf from New Beetle

I have tested 5 x E6600 so far with Commando on aircooling.
Please note that those Vcore were set in Bios not actually measured and room temp were between 18C--20C.
　　　Lot　　　　 1.350V-MAXClock　　1.600V-MAXClock　　　　x6-MAXFSB
　#1(L629A526)　　　　　3618　　　　　　　　3951　　　　　　　　 557
　#2(L629B948)　　　　　3246　　　　　　　　N/A　　　　　　　　 430
　#3(L633A808)　　　　　3690　　　　　　　　4095　　　　　　　　 528
　#4(L633A808-2)　　　3528　　　　　　　　3942　　　　　　　　 500
　#5(L628B127)　　　　　3609　　　　　　　　N/A　　　　　　　　 545

#1 & 5 were only CPU I could OC to over 5G and got U10s of modPI1M under LN2.

In order to OC over 5.3G under LN2, my guess is to find one of E6600 with:
1. capable to go over 3.6G at 1.35Vcore on air,
2. capable to reach over 560 of x6-MaxFSB

It is a kind of luck of draw, i guess:D

Hope you can find a good one:)

Interesting information :)

I can add L627B156 / 3.6GHz @1.30V-DMM OCCT stable @50C loaded / 5337MHz @1.85V-DMM CPU-Z suicide @-155C / 610FSB @-85C...
03-28-2007, 10:04 AM
fredyama

Quote:

Originally Posted by before

Interesting information :)

I can add L627B156 / 3.6GHz @1.30V-DMM OCCT stable @50C loaded / 5337MHz @1.85V-DMM CPU-Z suicide @-155C / 610FSB @-85C...

Thank you so much for your addition, before-san:toast:
03-28-2007, 11:04 AM
Flow

Quote:

Originally Posted by PRTeamJapan

>Flow: What clock does this cpu reachs watercooled @ 1,4V ... or how much does it need for 3,6ghz!

Additional inf from New Beetle

I have tested 5 x E6600 so far with Commando on aircooling.
Please note that those Vcore were set in Bios not actually measured and room temp were between 18C--20C.
　　　Lot　　　　 1.350V-MAXClock　　1.600V-MAXClock　　　　x6-MAXFSB
　#1(L629A526)　　　　　3618　　　　　　　　3951　　　　　　　　 557
　#2(L629B948)　　　　　3246　　　　　　　　N/A　　　　　　　　 430
　#3(L633A808)　　　　　3690　　　　　　　　4095　　　　　　　　 528
　#4(L633A808-2)　　　3528　　　　　　　　3942　　　　　　　　 500
　#5(L628B127)　　　　　3609　　　　　　　　N/A　　　　　　　　 545

#1 & 5 were only CPU I could OC to over 5G and got U10s of modPI1M under LN2.

In order to OC over 5.3G under LN2, my guess is to find one of E6600 with:
1. capable to go over 3.6G at 1.35Vcore on air,
2. capable to reach over 560 of x6-MaxFSB

It is a kind of luck of draw, i guess:D

Hope you can find a good one:)

Thank you! Thats a nice information ... I didn't think that your cpu just reaches 3,6ghz with 1,45V! The most important thing is propably the max FSB ...

My current E6600 reaches 3,6ghz with 1,31Vcore ... Next time I'll test this on a P5B for max fsb ...
03-28-2007, 06:07 PM
Eldonko

Great info TeamJapan, thanks for sharing and congrats!
03-28-2007, 07:33 PM
Kain XS

Long Time no see Team Japan in action.
Actually, is great to see this.
03-29-2007, 06:23 AM
fredyama

By New Beetle: additional results
Set up was almost identical to his previous 1M run except 1.79Vnb (Actual)

Max Clock with 1-core: 5339.85MHz (FSB593.32MHz x 9)　
2-click up from 2-core but modPI 1M was limited to the same clock speed as 2-core-- no improvement, unfortunately.

http://premium1.uploadit.org/fredyam...x-5339MHz-.png

modPI 32M: 9m56.625s @5255.9MHz (FSB584.0MHz x 9) Finally got U10m:)
http://homepage2.nifty.com/NewBeetle/32M_9m56s625.png

modPI 4M: 55.390s @5267.9MHz (FSB585.3MHz x 9)
http://homepage2.nifty.com/NewBeetle/4M_55s390.png

hexus pifast: 18.38s @5279.8MHz (FSB586.6MHz x 9)
http://premium1.uploadit.org/fredyam...ast_18s38T.png
03-29-2007, 06:30 AM
k|ngp|n

Awesome clocking :toast: :toast:
NEW BEETLE rules :D
03-29-2007, 06:38 AM
Viss

Very nice to see TJ results here again :toast: How much slower was 1M with 1 core and same clocks compared to 2?
03-29-2007, 10:05 AM
giraffen

Damm i like that setup :toast:

/Jesper

Show 100 post(s) from this thread on one page