Brisbane Testing in progress...

[breakfromyou]

ahh, so the first batch of 65nm AMD's clock bad?

[Nedjo]

OK!
Here is what we all been waiting for: First Brisbane overclocking scores on the Interent:



In the BIOS of this Foxconn mobo Vcore is set to 1.5V.
Corsair Dominator working on 1033 MHz@2.45V
This is the max OC on this Air cooling. I was hoping for 3.2 GHz, and went all the way to 1.7 V in the BIOS, but could not make it in to Windows. As a matter of fact anything beyond this frequency was not stable!

[Lightman]

Not bad for Air Colling .

Now how low you can go on voltage @ 3100MHz?
The other thing I'm interested in is mem bandwidth. Is it same as for RevF or is it better?

So far and for good work!

[STEvil]

But still there’s a question whom to trust: mobo BIOS, or CPU-z. For some reason, I’m on the CPU-Z side.

Digital Multimeter.

[Nedjo]

Concerning the temperatures, here’s what Core Temp (latest build that can recognize Rev. G CPU – kudos to The Coolest) has to say:

Idle:



Load (Orthos running):



Like I wrote, CPU is under BOX FX74 cooler.

[awdrifter]

If that temp is correct, you should be able to push it even higher. Too bad it maxes out at 3.1ghz.

[informal]

Kudos to the man who published the first OCing results for Brisbane on the internet!!!

And for the good air cooling,early Brisbane made a good show ,3.1Ghz air cooled is great!

[Nedjo]

And here are some benchmarks under 3.1 GHz!

[gn0s1s]

those are looking pretty nice! Could it not handle the 13x multi?

[Nedjo]

If that temp is correct, you should be able to push it even higher. Too bad it maxes out at 3.1ghz.

I remember the times when it was smarter to buy Newcastle instead Winchester ‘cos the newest CPU, made in the newest tech, did not clocked like its predecessor.

It’s good to see that’s not a case this time around. Also we shouldn’t loose from the sight the fact that there’s more to come in form of improving 65nm tech with bunch of interesting “spices”.

The most important thing about Brisbane is that is cooler (someone who’s building fanless machine could care about this), and that AMD can build much more of them making the space for even more price clashes.

Me personally expect that thanx to Brisbane, AMD will dramatically improve their positions on the notebook market.

[Nedjo]

those are looking pretty nice! Could it not handle the 13x multi?

It could, but I wanted to push Dominator more It works flawlessly in all the multi+HTT combinations that are resulting in 3.1 GHz

[Fred_Pohl]

Those temps are clearly way off but the difference from idle (~10C) to load (~32C) is probably correct. ~20C higher under load than idle would likely translate to ~30C idle, ~50C load given a typical ambient temp of ~21C.

[madgravity34]

Is your barrier at 3100 mhz an exact thing? Someone said that software issues could possibly be limiting the oc's? You should try around 10 - 50 mhz more and see if anything in there you can get stable. If you cant get anything stable at all or even remotely stable, wouldnt that be evidence of a bug somewhere other than the chip? Or maybe a limiter somewhere in the system? I have never really experienced a cpu that would clock right up to a number and wouldnt do anything higher at least for a few minutes!

[XFxGeforced]

Waiting for an update hopefully it hasn't hit a for sure wall yet.

[Nedjo]

In my book this has the same importance like overclocking:
Lowest stable voltage for the stock frequency (2.6 GHz):



I’m off to bed

[stealth17]

In my book this has the same importance like overclocking:
Lowest stable voltage for the stock frequency (2.6 GHz):



I’m off to bed

That's pretty sick but I won't even look at it until you get the DMM out and check the actual voltages. For all we know it could be 1.4v even though CPUz says otherwise.

[XFxGeforced]

In my book this has the same importance like overclocking:
Lowest stable voltage for the stock frequency (2.6 GHz):



I’m off to bed

no dont leave! more overclocking! loving those low volts too!

That's pretty sick but I won't even look at it until you get the DMM out and check the actual voltages. For all we know it could be 1.4v even though CPUz says otherwise.

eh, good point, but we can hope

[Apocolipse]

So is that super-pi score faster than a regular 90nm x2 at the same speed? Or is there no difference.

[XFxGeforced]

damnit he left. He should be required to stay till we're satisfied :p don't want to anger the man with the treasure though or he won't share.

[XFxGeforced]

No STATISTICALLY significant difference. There might be a few points in one direction or the other, but not more than 1-2%.

From the Japanese benches we saw earlier this week, it looks like there were minimal or no core enhancements in this shrink.

i wasnt really expecting much of a clock for clock performance gain, if any. Mainly interested in the lower voltage/overclocking potential.

[frankR]

I'd be satisfied if these could do 3.6 Ghz stable on air.

4 Ghz on chilly water might be asking for too much...

[NapalmV5]

And here are some benchmarks under 3.1 GHz!

not bad..

can anyone compare a conroe @ 3.1 to this?

[XFxGeforced]

I'd be satisfied if these could do 3.6 Ghz stable on air.

4 Ghz on chilly water might be asking for too much...

that'd be way too damn sexy

not bad..

can anyone compare a conroe @ 3.1 to this?

ill check that super pi score
*itd be easier if he did a 1m run, but i guess someone with a C2D could test it out, plz

[soulrider4ever]

If I get around to it, I'll run a 3.1c2d 8mb tomorrow morning.

[derektm]

Core will demolish it.