Athlon 64 Overclock on Air...

[DeathMonk]

Does anyone recall the highest they've seen on air (stable)?

[charlie]

About 2500ish, 3D mark stable....I think my fx-51 went 2560...

C

[DeathMonk]

How many volts? I'm at 2.4Ghz right now @1.6v... I haven't tried higher yet though..

[SmokeyTheBandit]

Pc-ice has got the frikkin highest Air OC 4 an A 64

A64 3000+ CAAOC 0352 XPMW

it can run @ 2753 mhz @ 1.78 V i believe :drool:

[Bravo]

Originally posted by SmokeyTheBandit
Pc-ice has got the frikkin highest Air OC 4 an A 64

A64 3000+ CAAOC 0352 XPMW

it can run @ 2753 mhz @ 1.78 V i believe :drool:

Running close to 0C air temps :)

[Maxvla]

Originally posted by Bravo
Running close to 0C air temps :)
uh.. how is 0c temp possible on air? any air cooled heatsink is going to be hotter than that unless he's using some kind of exotic cooling for the air hitting the heatsink... and i couldn't really call that air cooling.

[interman]

i think he meant the air he's blowing on the heatsink is 0c, not that the cpu is 0c.

[DeathMonk]

Well at 1.6v, (2.4Ghz) I am not completely stable. The next step up in the K8NNXP is 1.65v. Right now, if I sit in the bios it starts at 33'C and gradually goes up to around 40-41'C idle. If I use the Gigabyte utility that came with the mobo it sits around 33-34'C and doesn't move at idle. Which one is more right? Or should I just go for the average between?

Temps in consideration, do you think I have enough headroom left for 1.65v?

btw I have the zalman CNPS7000-Cu hsf

[sandman]

Man, my 3000+ must suck.

Needs 1.7V for 2.36ghz.

[DeathMonk]

On air?! If so, what's your idle temp?

[sandman]

No, it's on water.

Idle is about 31C usually.

[NeGe0]

I hit 2632 on my fx on air. Then I tried to run 01 and it crashed. I don't know what the highest I got it stable on. I know atleast 2540 is stable.

[luihed]

The Fx and 3000+ seems to clock higher than the 3200+ on air.... mine topped out at 2350mhz @ 1.7v on stock hsf, on the chiller, she does 2650mhz........

[sandman]

What vcore and cooling was it?

[TheDude]

Mine did 2530 3D stable on air 1.7 vcore 2.9 ram....2700 in MachII 3D Stable...screenshot at 2800....expect good results after volt mod is done.

[sandman]

jesus, mine is terrible.

Luihed what temp was your chiller at?

[DeathMonk]

Like I said earlier, i'm at 2.4ghz 1.6v... I'm more concerned about my temperature question (see post 8)..

I know this chip has great potential..

Also, does anyone know where I can find info on that mod for my zalman where you change the fan?

[luihed]

sandman, water temp was at -20c and mbm5 shows cpu temp at 3c...... once I run prime95 though it shoots up to 21c right away..... vcore is at 1.9v-1.95v......

[sandman]

thanks Luihed.

Deathmonk, You can't trust thos sensors, but if I was going to take a geuss I'd say you're fine as long as you're stable.

Within reason of course, if the sensor says you're at 75C, well you're definatly running hot, not sure about 75C, but hot.

[DeathMonk]

So... Last night I went home and set my fsb to 237, where it was completely stable at 1.6v. Trying to get more, I pushed the voltage to 1.65v in the bios (cpu-z read 1.696v). I then ran 3dmark01 and 03 a couple times to see if the extra voltage made it unstable. It was still completely stable..

But, whenever I go past 237 it becomes unstable (as with 1.6v).. It seems like all the way to 242fsb it does the same thing--drop to windows right before the loading screen of nature.

I tried a couple things that I could think of to no avail. Lowering the LDT to 2x didn't help, either did upping the VDDHT or VDDQ. THe only thing I haven't really tried yet was lowering the mem speed from 166 to the next lower value..

[benonmsn]

2530 for me, 25055 3dmark. stable. venus 12 heatsink, 1.6v. 98f idle, 117fafter 3 hours of prime

[DeathMonk]

That's a nice oc!

I lowered my mem speed to 150 and could get to 240fsb stable, still at 1.6v--I don't think cpu voltage is my problem right now because raising it has no effect. At 166mem, I was only doing a 3mhz overclock (237fsb).. why would lowering the mem to 190mhz make it stable to 240fsb?! At LDT 3x and mem 1:1 I could get 220mhz With the multi at 9.5 (LDT 2.5, 1:1) I could get around 235mhz out of the memory.. This is confusing stuff.

[Stang_Man]

Originally posted by Maxvla
uh.. how is 0c temp possible on air? any air cooled heatsink is going to be hotter than that unless he's using some kind of exotic cooling for the air hitting the heatsink... and i couldn't really call that air cooling.

why not? that's sooo easy. here in CT, the temperature outside is 15F. that is Below 32F/ 0C. Ambient temperature would be around -8C at such a temperature. processor wouldnt heat up as fast with that much cold air, and overclocking is easy.

2 weeks ago, we had -15 to -25F temperatures. that kind of temperature, easily matches the temperature of the head of a prometia or vapochill

:)

[some_young_guy]

i've been testing my A64 3200+ system at 252 fsb, no volt mods, @1.75 Vcore for a little while now.

It's been -30C to -40C, and 0% relative humidity all week here in Calgary. So I bought a 7 foot aluminum dryer duct, attatched one modified end to the window and the other end fit perfectly over the zalman flower heatsink. At full load the CPU hit 3 C max. But when idle, my Bios wouldn't read anything below zero for some strange reason.

I'm sure the readings would be lower if I brought the whole case outside, but i didn't want to f**k anything up...not to mention freeze to death.

[DrJay]

For what it's worth.........
2,430mhz @ 1.6v ( A64 3000+) Did 2,330 @ 1.55v

[saaya]

3000+=less cache=less heat=high oc on air...

plus if you buy a 3000+ you have a slightly higher chance to get the latest stepping...

[QuadDamage]

your temps will always go up in bios whether you're using air or prommie. For some reason bios heats up the chip.

Mine can do around 2.4Ghz on NF3 mobo and over 2.7Ghz on prommie but it was long time ago when there was on win based o/cing utility.

[saaya]

in bios your chip is close to 100% cpu load ;) there is no idle proggie like its implemented in windows xp that shuts down some parts of the cpu when not used.