P5WDH to evga 680i: Same CPU better OC + less volts?

[jeeka]

Hey guys,
Just wondering, IN GENERAL, did moving from a P5WDH to an evga 680i, with the same CPU yeild a "better" OC (higher OC/less volts) with the same CPU?

Reason being, I jumped on an e6700 (week 28) and it will do 3.65Ghz at 1.6v (365x10) on the P5WDH. It can do 390-400x9 with my old e6600.

I have a A1 D00 evga 680i and am anxious to see if this board will help get either CPU to a higher OC. It seems my e6600 will crap out at about 3.5Ghz to 3.55Ghz on the P5W with 1.6v. (1.6v is too high for my liking)

Is it possible, or regularly achieved, getting the same/better OC with less volts on the evga board, using carry over parts?

Thanks!
jeeka

[crazyea]

from my experience no. you will also lose performance moving to the evga.

[jeeka]

Ya I figured I might clock for clock, but if I can hit 3.7Ghz in the 680i as opposed to 3.5Ghz on the P5WDH at the same multi (using either the e6600 or 6700), then that clock for clock comparo becomes a non-issue since the P5WDH is not able to hit that high of a clock (assuming the 680i can)....

Either way, my wife is getting a killer machine, and I will be getting an even better setup once I figure out which CPU/MB Combo she is getting :) (e6600/e6700 with P5WDH/evga 680i)....

I'll fire it up and let you guys know all my results. I'm sure that my addiction to new parts might help someone in their quest for the 'perfect' machine.

Thanks!
jeeka.

[crazyea]

I went from the exact same set ups you are talking about here.

e6600@3.2 -1.36v (400x9) & P5WDH to the evga 680i A1 D0 and could not get any better (could only do the exact same oc).

Also even though I could tweak the memory more on the 680i (ie, lower timings, 1t), I could not match the performance of the 975x at a little looser timings.

The p5wdh is also passively cooled, while the 680i had that loud fan on the nb.

Needless to say the 680i is sitting in my closet and I happily went back to my p5w dh.

Good luck, maybe you'll have a better experience than me.

[jeeka]

Thanks for the well wishes.

I am thinking I can get a higher FSB at a lower multi (maybe 8x450-465).

My Ram (check sig): I was able to get it to run at 4-4-4-12@2.1v at 460/920 (rated for 400/800).

My thought is that I can overcome the loos(er) timings/performance when ramping up the FSB (when does the 1333 strap kick in on the 680i?) to 465x8 as opposed to roughly 390 (e6600 390x9) on the P5wDH. I have already shown that the CPU can hit 3.65Ghz all day long on the P5WDH, I just want more memory bandwidth (+100FSB if I can hit 465x8 on the 680i)

One more biggie: Does the 680i have the same issue as the P5WDH where if you try and boot at a lower multi than stock, initially it has to try and boot at the stock Multi then quickly shift down to the lower multi right away?


If that is the case, 400x9 for example, would boot at 400X10 (not happening), then multi down to 9x....that would really suck if that is the way it works on the 680i, too......

[crazyea]

One more biggie: Does the 680i have the same issue as the P5WDH where if you try and boot at a lower multi than stock, initially it has to try and boot at the stock Multi then quickly shift down to the lower multi right away?




Of course, this is the only way as the cpu has a locked multi.

[safan80]

Hey guys,
Just wondering, IN GENERAL, did moving from a P5WDH to an evga 680i, with the same CPU yeild a "better" OC (higher OC/less volts) with the same CPU?

Reason being, I jumped on an e6700 (week 28) and it will do 3.65Ghz at 1.6v (365x10) on the P5WDH. It seems my e6600 will crap out at about 3.5Ghz to 3.55Ghz on the P5W with 1.6v. (1.6v is too high for my liking)

Thanks!
jeeka

that's high vcore for the P5WDH I'm only using 1.52 on my 6700 style Xeon on my p5n32-e SLI plus..

I would suggest that you use water cooling on your NB & cpu if your not already.

[crazyea]

Limitation of the cpu, not the mb. I also have to crank 1.6v on both the 680i and the p5wdh to get 3.6.

[crazyea]

Here's a couple of links to some benchies I did while trying out these m/b's myself. All done at 400x8 1.36v 1:1 4-4-4-10 + ad CR1 for evga.

evga680i Sandra (http://www.pbase.com/eltonandersen/image/78037807/original.jpg)

P5W DH Sandra (http://www.pbase.com/eltonandersen/image/78037705/original.jpg)

Everest Comparison (http://www.pbase.com/eltonandersen/image/78037595/original.jpg)

[Daveb2012]

I think some 965 boards can out perform 680i with the same CPU.

[crazyea]

yes, it (680i) is known to be the slowest of the 3 chipsets. keep in mind that the variation between all 3 is very slight. The only bench that shows huge improvement is the everest photoworxx test.

[Solarfall]

I went from the exact same set ups you are talking about here.

e6600@3.2 -1.36v (400x9) & P5WDH to the evga 680i A1 D0 and could not get any better (could only do the exact same oc).

Also even though I could tweak the memory more on the 680i (ie, lower timings, 1t), I could not match the performance of the 975x at a little looser timings.

The p5wdh is also passively cooled, while the 680i had that loud fan on the nb.

Needless to say the 680i is sitting in my closet and I happily went back to my p5w dh.

Good luck, maybe you'll have a better experience than me.


when was this ??? the new evga p27 bios kicks some serious arse.. now i'm able to reach +4ghz speed with my conroe with p24 i was stuck to 3.7ghz

[crazyea]

certainly this is not typical and I wouldn't advocate buying a new board based on a hypothetical o/c. There are as many people on the evga forums who have had no benefit compared to those who can suddenly o/c more. Also, I wouldn't compare a bios with a new strap lowering performance a better o/c anyway.

[safan80]

Limitation of the cpu, not the mb. I also have to crank 1.6v on both the 680i and the p5wdh to get 3.6.

What is the batch code of your cpu?

[crazyea]

Here's a pic, Pre-Lapped. (http://www.pbase.com/eltonandersen/image/73866313)

[safan80]

it's a 32B.. I still think your using too much vcore.

[crazyea]

The pc won't boot at 3.6 any lower than that. I use 1.36 at 3.2 and, for me anyway, the benefits of using a lot more for a few extra mhz is not worth it.

[safan80]

The pc won't boot at 3.6 any lower than that. I use 1.36 at 3.2 and, for me anyway, the benefits of using a lot more for a few extra mhz is not worth it.

did you try to OC before you lapped it?

[crazyea]

lol, yes.

I know it's an honest question, but I kinda know what I am doing.

[safan80]

lol well your 32B reminds me of my 42A :( if I were you just hold off until the newer 1333 fsb chips come and then worry about your motherboard.

[crazyea]

Right now I have a x2 4200+ EE in my htpc, so I might even splurge on a new cpu in abit and upgrade that. We'll see. The new 6420's are looking to be good clockers. Maybe I'll try a xeon. A real dilemma for sure. :-)

[Falkentyne]

Of course, this is the only way as the cpu has a locked multi.

Nope, this happens with X6800 CPU's too on the DH.
No one here has determined (after months? Shame on you guys!) if this is a BIOS bug (relating to locked chips -->unlocked) that is not supposed to be affecting unlocked chips, or if this is a 'feature'.

System tries to POST with the X6800 chip at the default multiplier at the FSB you chose. If you chose 430 * 8, it will try to POST at 430 * 11. If the chip can't handle that speed--no POST. Do other boards have this problem or is it asus only bug? No one answered that after 6 months...

[crazyea]

Thats a different issue altogether as he was referring to an e66/6700. Any motherboard the allows downward multi's on a locked cpu would have to use this method.

[jeeka]

Exactly what I was trying to ask Falkentyne, thanks for the clarification to what I thought was an already clear question :toast:

That's what I was trying to figure out myself.

Also, with the p27 BIOS, are we sure that looser internal timings have been introduced to help with higher FSB's? It makes sense, but all we get is the "better OC'ing capability" line.... Us OC'ers want to know specifics evga!