ASUS P5N-E SLI 650i Thread

[Christiaan]

Originally Posted by Ravin

So you have two options.... Running TRUE 1:1 will put your RAM @ DDR rate of 700MHz, or run QDR 1:1, which clocks your RAM @ DDR of 1400MHz.

For me, downclocking my RAM from DDR1000 to DDR700 is not a good option- I take a huge performance hit. DDR800 would be acceptable for me if I could get 3-3-3-6 1T, but I cant post past 375MHz FSB to get 400fsb/1600QDR/800DDR

Hehe, ahh, but you see, I have DDR2 667 RAM.
So I am overclocking them.

[Ravin]

Originally Posted by Christiaan

Hehe, ahh, but you see, I have DDR2 667 RAM.
So I am overclocking them.

Yup. This is true. I'd say go with the quad then. IMO a 25% OC is not too shabby- especially if you can undervolt to run the CPU cooler. My VID is 1.2875V, currently running 1.250V (and plan to drop at the next reboot). My temps are an insane low near 50C load on the stock intel cooler.

You will always be able to upgrade to P35/X38/X48 board to get a higher OC once the quad starts to lag behind in performance. Probably by then Intel will be 2 archetectures down the road.

[TriBeCa]

mosfet sinks in place, i'm back running my sig OC. dunno if it's stable yet, but i haven't seen any random reboots.

also...something freakish has happened. Installing the vdimm mosfet sinks required removing the CPU cooler. at 1.56vcore (1.48 under load) my temps appear to be 8-10C cooler than they were before? ***? and my last cooler seating was like 2C over ambients max stock idle. can anyone explain this? AS5 hasn't even cured yet.... we'll see how the temps are in a few hours, but....

[Ravin]

Originally Posted by TriBeCa

mosfet sinks in place, i'm back running my sig OC. dunno if it's stable yet, but i haven't seen any random reboots.

also...something freakish has happened. Installing the vdimm mosfet sinks required removing the CPU cooler. at 1.56vcore (1.48 under load) my temps appear to be 8-10C cooler than they were before? ***? and my last cooler seating was like 2C over ambients max stock idle. can anyone explain this? AS5 hasn't even cured yet.... we'll see how the temps are in a few hours, but....

Maybe the extra heat dissapation on the MOSFETS helps out the CPU. They are pretty close together, and the MOSFETS run hella hot when the Vcore is cranked. That or you just got a really good mount on your HS.

Edit: now testing 1.235Vcore.....

[TriBeCa]

i thought of both of those...but I notice now that the MB temp is like 10C cooler, as is the GPU???

And I really, really doubt it's the HS seating, as the last setup idle on stock at like 2C over ambients. Not really possible to get much lower than that, so I assume the last seat was good....

[Ravin]

Originally Posted by TriBeCa

i thought of both of those...but I notice now that the MB temp is like 10C cooler, as is the GPU???

And I really, really doubt it's the HS seating, as the last setup idle on stock at like 2C over ambients. Not really possible to get much lower than that, so I assume the last seat was good....

very interesting indeed. Room temp the same? I'd say that it's likely that reducing strain on the MOSFETs by cooling is giving you cleaner power all around, reducing the load on other components. But that is guestimation from experience in industrial electric supply.

For example, You have a motor rated to run 480V/10A @70F ambient, but ambients can climb to 120F. This will increase the resistance of the motor coils, thus reducing current @ the motor. All things being equal, you are still using 480V, and still pulling 4800W, just the resistance and current are slightly different. The electronics inside the motor control panel will also see an increase in thermal output, due to the increased resistance. I've seen many starter switches burn out in just this fashion.

So with your setup that you are seeing the opposite effect using the same principle.

50 minutes strong @ 1.235Vcore

[TriBeCa]

makes sense--i forgot about that increased resistance with increased heat thing, though I saw a cool demo of it recently involving liquid nitrogen and an iron ring.

i'd love to see someone else pushing >1.5vcore try cooling their mosfets and see if they get a similar effect....

if this OC is stable I'm gonna try pushing my OC up further. I'll FSB overclock again, see if I can push past my prior wall of 440 FSB, and then try to get stable a little further up.

either way i'm loving these temps

[Ravin]

Originally Posted by TriBeCa

makes sense--i forgot about that increased resistance with increased heat thing, though I saw a cool demo of it recently involving liquid nitrogen and an iron ring.

i'd love to see someone else pushing >1.5vcore try cooling their mosfets and see if they get a similar effect....

if this OC is stable I'm gonna try pushing my OC up further. I'll FSB overclock again, see if I can push past my prior wall of 440 FSB, and then try to get stable a little further up.

either way i'm loving these temps

Did you put sinks on all the MOSFETS or just the ones near the LPT/PS2 ports?

[Kopi]

Ram my NB at 1.5 all the time, never touched 40c.

Its orthos stable and all, just gaming takes it down.

Tried a different PSU and CPU and Ram and GPU

[TriBeCa]

all the mosfets except the two or three that are down by the front chassis fan connector. I have no idea what those ones are for, but the sinks i have wouldn't fit in there anyway as they're too close to other board components.

i have 5 sinks on the 8 mosfets between the CPU socket and the I/O port panel, and 2 sinks on the 3 mosfets above the CPU socket (next to the where the PSU is in a normal case). those little anodized aluminum zalman vram sinks i'm using have a pretty impressive amount of surface area for their size.

later tonight or tommorow i can post pics of the sinks on the CPU mosfets if you want. can't do the vdimm ones without removing the cpu cooler again, so i'm not gonna do that