3700+ SD at 3.4ghz PR rating

[stealth17]

Would a 3700+ SD that runs 2.2ghz stock, if you overclock it to 3.4ghz, would the PR be about 5712, but a bit more because of a htt increase (300~htt) making it over a 6000?

2200mhz*1.68=3700 (or there abouts)

so 3400mhz*1.68=5712 (+ a few for the huge htt increase)

and i figuring this right?

man, that is huge, 6ghz! woot!....that a good or better than a p4 at 6ghz :woot:

[Ackbar]

Would a 3700+ SD that runs 2.2ghz stock, if you overclock it to 3.4ghz, would the PR be about 5712, but a bit more because of a htt increase (300~htt) making it over a 6000?

2200mhz*1.68=3700 (or there abouts)

so 3400mhz*1.68=5712 (+ a few for the huge htt increase)

and i figuring this right?

man, that is huge, 6ghz! woot!....that a good or better than a p4 at 6ghz :woot:

PR is kinda m00t when the corresponding CPU (intel) have not been created. It would be like extrapolating data beyond the dataset and assuming some simple linear or quadratic relationship without knowing the true underlying function.

[frostedflakes]

What I'd like to know is where the SD at 3.4GHz is? On single-phase or cascade?

[Zebo]

Your PR is whacked

it's
1.5 for newcastle 754
1.53 for CH 754
1.55 for 939 512kb chips
1.57 for 939 1mb chips

so in other words = 5338Mhz PR

[Zebo]

What I'd like to know is where the SD at 3.4GHz is? On single-phase or cascade?

there one @ 3500mhz on waterchill someplace twards end of this thread
http://www.xtremesystems.org/forums/showthread.php?t=59521

[stealth17]

You're PR is whacked

it's
1.5 for newcastle 754
1.53 for CH 754
1.55 for 939 512kb chips
1.57 for 939 1mb chips

so in other words = 5338Mhz PR

how do you figure?

a 3700+ SD runs 2.2ghz stock...so 3700/2200mhz=1.68....

i dont see where i messed up :confused:

What I'd like to know is where the SD at 3.4GHz is? On single-phase or cascade?

well its mostly speculation atm...people are hitting 2.8 on a ZP-120 air cooler so im figuring if im lucky, my -60c single-phase should be able to hit close to 3.4ghz +/- a bit....

[afireinside]

-60C? Idle I assume? What does it hold under load?

[Zebo]

I figure because I've entered about 4000 figues from reviews around the net into an excel file and came up with real PR's, Not AMD's fusked up numbers.

You know a 3400NC beats a 3500 in almost every benchmark? Hows that possible? AMD pulls the numbers from you know where that's how.

[Grov]

Your PR is whacked

it's
1.5 for newcastle 754
1.53 for CH 754
1.55 for 939 512kb chips
1.57 for 939 1mb chips

so in other words = 5338Mhz PR

But a 754 Claw, is faster than any 939 512kb chip still.

[stealth17]

-60C? Idle I assume? What does it hold under load?

nope...holds it load....got a hell of a compressor! :banana:

[Zebo]

But a 754 Claw, is faster than any 939 512kb chip still.
:confused:
Maybe a 3700CH vs 3000winch claw is faster. You got to compare at same speeds though


Nope. eg. a 3700 (2.4) does'nt win a single benchmark over 3800 (2.4)
http://www.behardware.com/articles/531/page1.html

[overclock shot]

754 Claws faster, sweet, how high have those been taken on air?

[blinky]

nope...holds it load....got a hell of a compressor! :banana:i can garuntee you its not -60c load under single stage, unless you have like no capacity and are running in deep vacuum, are are autocascading.

[stealth17]

i can garuntee you its not -60c load under single stage, unless you have like no capacity and are running in deep vacuum, are are autocascading.

i dunno..i didnt make it...its atleast -50c, thats what my thermometer maxes at...its got some special mixture in it too

[Grov]

:confused:
Maybe a 3700CH vs 3000winch claw is faster. You got to compare at same speeds though


Nope. eg. a 3700 (2.4) does'nt win a single benchmark over 3800 (2.4)
http://www.behardware.com/articles/531/page1.html

Don't think so, it wins plenty, thats just one link. Check out hardocp, 3700 beats 3800 in like half of benchies frok last year.

A 2.2Ghz (3400) claw, would be faster than a 2.2 (3500) Winchester in lots of apps.

Everyone knows it depends on the apps. Everyone seems to think 754 sucks, but it's just as fast as 939 really.

[IvanAndreevich]

Don't start another 754 vs 939 war - it's been done already.

[Saruji]

Don't think so, it wins plenty, thats just one link. Check out hardocp, 3700 beats 3800 in like half of benchies frok last year.

A 2.2Ghz (3400) claw, would be faster than a 2.2 (3500) Winchester in lots of apps.

Everyone knows it depends on the apps. Everyone seems to think 754 sucks, but it's just as fast as 939 really.

Yea umm well I have yet to see some benchies from anyone that said at clock for clock the 754 would beat a 939. Not to mention most 754 claws will max out faster then a 939 chip. I've just traded in my nf3 dfi 3400 claw that did 2.5 stable for a 300 winni that cost me about 40 dollars less and is currently running 2.6 @ 1.55 100% stable, and i'm just getting warmed up, waiting for my wc setup and i'll have it running prime ready at 2.7, my max benchie was a 2.75 so far. So yea, i'd say the 939's have a bigger advantage over the 754 no matter which way you look at it. Anandtech, tomshardware, devhardware all have posted the same results, even if you did dig up some benchie that read otherwise i'd be speculating because so many big names in the business who's work I trust more are saying otherwise.

[[XC] leviathan18]

if venice clock for clock is faster than newcastle and winnies in socket 939 i think they will be faster than socket 754

[The Mofo]

Well my 754 Claw systems does beat out my winnie system, equally clocked in D2Ol results. I couldnt believe it myself, but the stats dont lie... It had me scratching my head for a few weeks.

http://www.venomousmofo.com/mofo04/winnie.jpg

[stealth17]

Well my 754 Claw systems does beat out my winnie system, equally clocked in D2Ol results. I couldnt believe it myself, but the stats dont lie... It had me scratching my head for a few weeks.

http://www.venomousmofo.com/mofo04/winnie.jpg

thus read Saruji's post

Not to mention most 754 claws will max out faster then a 939 chip.

max out both chips and if the winnie is even a half good stepping it will pull ahead...

the venices have seemed to make more performance at a lower speed too...not to mention san diego vs. clawhammer

crunching is purly cpu speed too...heck a friggin Mac can do that :rolleyes:
benchy the 754 claw against a 939 winnie in something that uses memory...with some tccd that does 7500mb/s+ sandra, the winnie would shine due to dual channel

[Hombre]

well its mostly speculation atm...people are hitting 2.8 on a ZP-120 air cooler so im figuring if im lucky, my -60c single-phase should be able to hit close to 3.4ghz +/- a bit....

As other said above, there's no chance that you're running -60 load with SPC (unless you run one hell of a vacuum), maybe closer to the lines of -45 load. If you'll get lucky, you can hit 3.4 screenshot (with a really good chip obviously).

[perry_78]

Yes, -60C load does sound a tad optimistic, though it's still better than a kiddo telling me he has -350C cooling :rolleyes: (physics teacher sticks to point nothing goes below 0K, though some here disagree)

3,4ghz might be possible, look at Hardcoreclockers FX55, but luck would indeed play a large role regardless of your fantabulous cooling ;)

[stealth17]

well i havent recieved the unit yet but ill let you know once i do.

i just want atleast 3.2 24/7 100% stable...i cant stand an unstable system...crashing pisses me off