• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

The Super Pi Mod Score Database

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Celeron 2.4 - 122s

Wow, compared to my E6400@3.2, assuming performance on super pi scales linearly with clock speed on specific chips (poor assumption maybe), that means my celeron takes 5x longer to calculate pi, clock for clock.

That amazes me. My celeron would have to be clocked to over 10GHz to match the stock performance of my Conroe, or conversely, my Conroe could be clocked at 500MHz, and still out perform my 2.4 GHz Celeron.
 
Compiled a spreadsheet with the results so far. 54 results, some of which only had times to the second, so they may be a little inaccurate.

Yoxxy owns the speed crown by a fair margin, 1.602s. There isn't even close to that big of a gap between any two times until you get to 30th on the list. Yoxxy is followed by Verse914, Gary Key and garkon8.
Sorted by time

I also multiplied the Clock Frequency by the time to complete. This is sort of a clock per clock indicator. It basically shows how many processor cycles it took to complete the 1M task. Yoxxy also came out on top of this one, but by only the slimmest of margins. The 4MB cache chips owned this category. Yoxxy was followed by gersson, Yreka and lopri. Pentium class machines performed abysmally in this metric.
Ranked by Cycles

The index column is just the Cycles category normalized to Yoxxy's class leading standard. For example, a 1.2 here would mean your processor takes 1.2 cycles to do what yoxxy's did in 1. Same thing as the Clock cycles, but the magnitude of the number is more meaningful in relation to the other users' results.


 
Damn, I wouldn't have been able to come up with a chart like yours!

Amazing job Gerwen. Maybe if more people post, it can be stickied haha. Props! lol.
 
17.125

E6400 at 3.4ghz
2gb OCZ ddr-800 @ ddr-850 4-4-4-15
Asus P5B-E rev 1.02g
stock HSF


Wow! I see the 4mb cache of the E6600-E6800 really make a difference in SuperPI

My chip seems to have a bit left, but I'm not going to try it until I get better cooling.
At 3.4ghz on the stock HSF this baby runs HOT!!! It hits 70c:shocked: under full load
 
Gerwen, baby,
i love your work, but can i request Motherboard type and memory speed, brand in there as well please,
as i'm (Proud celery 2Gig, GForce440 owner) thinking of upgrading (WHY! i hear you all ask (well . . . 1 FPS C of Ridick)

also; anybody is welcome to advise me, add YOUR 2 cents NOW!
especially on what memory they would get to max SUPER Pi

$100 Biostar TForce P965PT
(589fsb for 100 bucks ain't bad http://www.xtremesystems.org/forums/showthread.php?t=110106&page=10 )
$130 E4300
cheap memory to fry it just for fun
http://www.techbuy.com.au/search.asp?wh...%240.00&srchNumRecs=100&rbSrchType=ALL
who says what?


(also i suspect the 2.4G celery is 82seconds ie 1 22 )


 
If you want cheap memory, search for 2gb ram on ebay selling for under $150. That's what I did.
Overclocking is a crapshoot. Sometimes you win big and sometimes you can end up like me - getting an S3 dummy that has FSB holes.
I would highly recommend an S3 however because the overwhelming majority have great success.
 
16.484s on my
E6400 @ 3.6 GHz (450*8) 1.55v
DDR800 @ 900 4-4-4-10
Actually got 16.251 but didn't get a screenshot because I was being stupid. Duh

17.931 on my current config (see my sig)
::EDIT - I'm back to my 3.6GHz Config
 
14.797s on E6600-3429MHz(381x9), Asus P5WDH. That is for now. Dual Prime95 stable.

I wish I could go higher. Temps and Orthos failures are problems.
 
Originally posted by: PooBeetle
Gerwen, baby,
i love your work, but can i request Motherboard type and memory speed, brand in there as well please,

(also i suspect the 2.4G celery is 82seconds ie 1 22 )
Click to expand...

I'll work on getting the Mobo and mem info. It's tough since it's a moving target sometimes. People will post 'specs in sig' and sometimes specs change without updating the scores. Also, there's some scores without those specs.

The 2.4G celery is definitely 122s, or 2m 2s, but it's my work machine, so i couldn't shut down some things i had on the go to run super-pi. Regardless, it's likely an underperformer in the best of conditions.

I hope to keep updating as long as folks keep posting. My request is that the significant specs be in the body of the post rather than the sig, ie:
CPU:
Clock Speed:
Motherboard:
Ram brand, type, speed, and amount:
Cooling:
and maybe even whether or not it's orthos stable.

**edit** updated to add mobo, ram and HSF info, and somehow i missed Benchzowner's results until now

Updated lists:
Sorted by time
Sorted by cycles


Anyone know how to conveniently post a table in the forum? It would be preferrable to images.
 
CPU: Sempron 3100+
Clock Speed: 2.25ghz
Motherboard: ECS NF3
Ram brand, type, speed, and amount: 1GB Corsair Value PC3200
Cooling: AMD Stock heatsink

40.343 seconds 🙂

...to be updated with C2D results in the coming week
 
I second the bump for sticky.
But, gerwen could you refresh your spreadsheet. My times sound better in the 12s.😉
 
to be meet the "official" post requirements, a post requires a single screenshot with multiple cpu-z showing every tab and super pi mod1.5 with the checksum and OK box. posting screenshots is not my thing, so i hope we can keep it friendly.

my first run better than 34s.

33.937s
checksum: 610B9A22

AMD Opteron 165 - 2700 MHz
Asus A8N32-SLI Deluxe
Corsair TWINX1024-3200C2PT 2x1GB - 199 MHz
 
Well, despite the fact nobody is providing screenshots, I think it's safe to assume that nobody would be making up numbers just for the hell of it. It wouldn't make much sense.

This topic is for reference purposes so that we can come up with an enormous pool of data from which people can reference for their own overclocking efforts.

However, it's up to everyone else.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top