TaPaKaH breaks SuperPi 1M record to 7.140 seconds!

[Andreas G]

Over the last week the SuperPi 1M world records has been beaten numerous times by onepagebook, duck, coolaler, and a team consisting of joe_cool, no_name and predator. In the end though, it was Latvian overclocker TaPaKaH that came out on top. Despite a frequency disadvantage, duck's Intel Core 2 Duo E8500 was operating at 6544MHz when reaching 7.172 seconds, TaPaKaH was able to beat him by 32 milliseconds. The new world record is 7.140 seconds.

TaPaKah used an Intel Core 2 Duo E8600 overclocked to 6376MHz, almost 200MHz lower than the previous record. A bit surprisingly, he also had a lower system bus; 637.6 vs. 688.9MHz. What makes TaPaKaH's achievement even more amazing is that his score is with both cores enabled, duck had only one core enabled.

He had a much higher memory frequency though, and that may very well be what tipped the scales. His Kingston HyperX 14400 DDR3 modules were working at 956MHz, while duck's TeamXtreem 9600 DDR3 modules were only running with 1:1 divider, I.e. 688 MHz. duck did have tighter timings, but overall we would estimate that the higher frequency made up for that.

Click to enlarge

[KTE]

Won't be up long, as he knows, there's already a faster one the previous holder didn't upload yet.

EDIT: actually not one, quite a few more.

[Delph1]

He still holds the record, but it's a new one.

I'm not going to debate the authenticity of TaPaKaH's results here. He's innocent until proved otherwise.

//Andreas
_________________
"I'm sort of a postmodern vegetarian, I eat meat ironically"
"You are our Swedish superhero "

Forget everything, knowledge is your enemy, it's only going to make you second-guess yourself."

[KTE]

Authenticity? Yeah on XS he's doubted but I don't really bother into it.

He knows duck has a faster score below .125 anyway

[nvidiaamd]

What method do these Overclockers use to disable a core?

Good to know
_________________
E6750 @ 3500 MHz 24/7 ~ P35-DS4 rev2 ~ 2GB PC6400 Cas4 ~ 500W FSP Bluestorm II

[KTE]

You can disable it from the BIOS or through Windows and CPU-Z usually.

Windows XP method is under Start> Run > type: msconfig > Enter > under "Boot.ini" tab > Advanced Options > /NUMPROC > <configure as appropriate>

That sets the number of cores the OS will detect on the next boot.

BTW, a side question... what FSB can you achieve max and stable on your DS4?
I have that board lying around untested, about to sell, hence.

Damn, maybe he can get 1M: http://valid.x86-secret.com/show_oc.php?id=399936

[nvidiaamd]

My Highest stable FSB is 475 but I have an E6750 that might be limiting my Mobo, that is a Rev 2.0.

Feel free to ask more Questions and ill do my best to answer them

/ nvidiaamd
_________________
E6750 @ 3500 MHz 24/7 ~ P35-DS4 rev2 ~ 2GB PC6400 Cas4 ~ 500W FSP Bluestorm II

[KTE]

Thanks. Yeah, a lot of E6750 can't do above the 475FSB region for some reason, many complained about this. My last one had v.similar limits but it needed 1.392v for 3400MHz properly stable which ran too hot for my liking, so had to drop to 3.3G region.

I have had a rev.1.1 of the MB not long ago though and it has done 480FSB stable using a Q6600 G0 although a Q6600 B3 could not get above 425FSB stable on the same MB. I think max unstable but benchable with the G0 was 490FSB. Max MHz was near 4.25G on air.

[Mean Machine]

P35 motherboards are generally limited by the CPU, not vice versa.

Besides, only select boards have the option to disable cores in BIOS.
_________________
Man, or woman, or both?
Grandis spiritus, sanctus diavolos

[super dank]

One of my cores seemed to disable itself once when I was over-volting

[Delph1]

[super dank]

No, I'm not

One of my E6600 cores was greyed out in Thermal Analysis Tool, I lowered the vcore and it was ok

[Mean Machine]

That was a new one. Sure you cooled that core enough, so it didn't overheat compared to the other one?
_________________
Man, or woman, or both?
Grandis spiritus, sanctus diavolos

[super dank]

I believe that was the problem, it happened before I built my water chiller, when I had only the radiators. The temps were quite high

[KTE]

Someone on another forum complaining of 86C stock temps had only 1core showing up with a new Phenom 9950 too and I had no idea what it was. We told him to check volts, etc. He came back and said he redid the TIM/HSF, changed volts from 1.38v to 1.3v which the MB was overvolting by stock, and now has 52C load temps and all 4cores showing, so it might be something related

I've personally never seen/heard it before apart from that example this last week and now super dank.

[KTE]

What I meant... http://www.ocxtreme.org/forumenus/showpost.php?p=51132&postcount=1