SuperPI 1M Performance to MHz Ratio

[Shimmishim]

Link to original thread.

Text:

Here are the rules.
1. All scores submitted must have a CPU clock of at least 2.4GHz.
2. Score must be accompanied by a Screenshot with CPUZ (main window) and A64 Tweaker open.
3. Super Pi Mod 1.4 MUST be used. The older verions and SEE2 / SEE3 versions will NOT be accepted.
4. Only 1 score, being the highest of your scores, will be added to the listing.
5. Post in the according format. Scores not posted in this format will NOT be posted. " 41.044 --- dinos22 ------------- Opteron 144 "
5a. The format is as follows. Time (to 3 decimal places) three dashes "---" Hyperlink to SP1M time. appx 10 dashes "----------" and finally your CPU and if it is up for question, the core type (IE. FX55 SD or FX55 Claw)

Here is the Formula
Your Time in Seconds * Your Mhz (please use as many decimal places of accuracy as possible) / 2000 = Your score

Example
27.703 * 2951.7 / 2000 = 40.88547255

Lower scores are better!

We'll divide this up into categories

1. A64 single core
2. A64 dual core
3. Pentium single core
a. netburst
b. dothan / yohan
4. Pentium dual core


Most important thing to remember is... POST YOUR SCREENSHOTS!!!

Example of what to post in your screenshot

39.535 --- funkflix----------- Opteron 146

 

[stevty2889]

Why in the world would you use the (time x mhz) /2000 rather than just the time?? You won't be able to compare Intel and AMD scores your way, beacuse Intel uses higher clock speeds, so the Intel scores will always be scewed higher, even if the AMD chips get better times..

 

[woog315]

Originally posted by: stevty2889
Why in the world would you use the (time x mhz) /2000 rather than just the time?? You won't be able to compare Intel and AMD scores your way, beacuse Intel uses higher clock speeds, so the Intel scores will always be scewed higher, even if the AMD chips get better times..

and like he said in the post, lower scores are better

 

[Shimmishim]

Originally posted by: stevty2889
Why in the world would you use the (time x mhz) /2000 rather than just the time?? You won't be able to compare Intel and AMD scores your way, beacuse Intel uses higher clock speeds, so the Intel scores will always be scewed higher, even if the AMD chips get better times..

that's why i said categories...

cuz you're absolutely right that the A64's would dominate most likely...

then again dothans are pretty good at superpi as well.


anyway... check post #1... i said i'd divide up things into categories

 

[stevty2889]

41.2128-----stevty2889---------Dothan 725
43.872444-----stevty2889.---------X2 4200+
69.4925025-----stevty2889-----------Pentium 4 540 ES

 

[RichUK]

I?m in... I recently purchased a 148 opty from lopri (thanks mate) which is capable of 3Ghz on 1.48 (proven), matched up with my TCCD should get me 28s?s ish, although I will push for <28's (wishfull thinking maybe). (im at 2.9Ghz @ 1.45 on my board at the moment, until i get a proper working BIOS. However im looking to get the eXpert soon).

Results pending.....

 

[Avalon]

43.358 --- Avalon ---------- Opteron 146

Running a cabbage at the moment. First time using A64 tweaker. I went to change a setting and my computer hard locked, so maybe I should leave it alone. Anyway, I could probably get a much better score, but I'm not done overclocking this chip and I'm too lazy to do this again

Well, I figured out why A64 Tweaker crashed. I went to change the memclock frequency but didn't change the refresh rate. Got it working now. So this program is considered safe?

 

[GuitarDaddy]

Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

 

[Shimmishim]

Originally posted by: GuitarDaddy
Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

no no no.

i didn't come up with the formula.

and yes the formula is biased towards lower clockspeeds slightly.


i thought this would be fun for AT to do...

but doesn't look like it's very popular so far.

 

[Keysplayr]

Originally posted by: GuitarDaddy
Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

LMAO!!! Why do some of you guys talk like comic book characters? "You can try to stop me Cabbage Man, but your "feeble attempts" are futile."

LOL. sorry for the OT. Just struck me funny.

 

[TecHNooB]

Originally posted by: keysplayr2003

Originally posted by: GuitarDaddy
Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

LMAO!!! Why do some of you guys talk like comic book characters? "You can try to stop me Cabbage Man, but your "feeble attempts" are futile."

LOL. sorry for the OT. Just struck me funny.

wtf?

 

[Elfear]

Originally posted by: keysplayr2003

LMAO!!! Why do some of you guys talk like comic book characters? "You can try to stop me Cabbage Man, but your "feeble attempts" are futile."

LOL. sorry for the OT. Just struck me funny.

C'mon keys. Some people actually use words that are longer than 4 letters.

 

[GuitarDaddy]

Originally posted by: keysplayr2003

Originally posted by: GuitarDaddy
Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

LMAO!!! Why do some of you guys talk like comic book characters? "You can try to stop me Cabbage Man, but your "feeble attempts" are futile."

LOL. sorry for the OT. Just struck me funny.

Whats up doc! which way de go! which way de go!

Sorry, us "old guys" use real words from the Websters not the Urban dictionary

I guess I should have said, "That formula is wiggidy whack!"

 

[GuitarDaddy]

Opps, double post

 

[GuitarDaddy]

Originally posted by: GuitarDaddy

Originally posted by: Shimmishim
no no no.

i didn't come up with the formula.

and yes the formula is biased towards lower clockspeeds slightly.


i thought this would be fun for AT to do...

but doesn't look like it's very popular so far.

Slightly is the understatement of the year, If you want to know what your 3.8 pentium would score running 2.0ghz this is perfect

 

[stevty2889]

Originally posted by: Shimmishim

Originally posted by: GuitarDaddy
Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

no no no.

i didn't come up with the formula.

and yes the formula is biased towards lower clockspeeds slightly.


i thought this would be fun for AT to do...

but doesn't look like it's very popular so far.

I think it's mainly because we like to be able to compare things directly. With this formula, you can only compare a cpu to the same type of CPU. A64's and Pentium-M's will always have the lower score(and lower score is better) compared to Pentium-4's, even in the case where the P4 might have the better time, although usualy the SSE3 patched version makes the biggest differance for Prescotts, and it does make a huge differance.

 

[Avalon]

Originally posted by: GuitarDaddy

Originally posted by: keysplayr2003

Originally posted by: GuitarDaddy
Time*mhz/2000 makes no sense

It looks like a feeble attempt to adjust all results to stock speed, if so the formula should be

Time*(o/c's mhz)/(stock mhz)

LMAO!!! Why do some of you guys talk like comic book characters? "You can try to stop me Cabbage Man, but your "feeble attempts" are futile."

LOL. sorry for the OT. Just struck me funny.

Whats up doc! which way de go! which way de go!

Sorry, us "old guys" use real words from the Websters not the Urban dictionary

I guess I should have said, "That formula is wiggidy whack!"

Best post of the new year

 

[cmrmrc]

65.457---------AXP 2800+ (2.078ghz)

 

[Alexstarfire]

Well, I'll post mine but I don't expect you to put it up since I don't have the A64 Tweaker program open in it. Did something like this over at Guru3D and is the only reason why I'm posting it, because I don't feel like running it again. Well, I do, but not now.

43.3205472---http://img288.imageshack.us/my.php?image=superpi7zm.png]Alexstarfire[/url]----------A64 3200+ Venice

When I do this again I'm going to beat you. I SWEAR. LOL, Dothan's will probably have the best ratios.