x2 3600+ Brisbane hits 3.1 GHz

[BaronMatrix]

For those who were wondering the folks at AMDZone managed to get their 1.9GHz 3600+ to run stably at 3.1GHz with 1.48V and the retail HSF.

Linkage!

Looking at this result makes me think that Barcelona will do 80% OC. I mean Brisbane is just a shrink where Kuma will be from the ground up 65nm with all of that DSL stuff showing off it's ability.

We'll see though. Since AMD did mention 1.8x faster per dual core, they obviously have the mobile version of Barcelona up and running to compare directly to Windsor.

AMD hasn't been talking a lot about mobile lately but the last word was that mobile and server would hit at the same time.

I am actually of the opinion that mobile should hit first since it is said that mobile revenue will overtake desktop this year.

 

[qcmadness]

For those who were wondering the folks at AMDZone managed to get their 1.9GHz 3600+ to run stably at 3.1GHz with 1.48V and the retail HSF.

Linkage!

Looking at this result makes me think that Barcelona will do 80% OC. I mean Brisbane is just a shrink where Kuma will be from the ground up 65nm with all of that DSL stuff showing off it's ability.

We'll see though. Since AMD did mention 1.8x faster per dual core, they obviously have the mobile version of Barcelona up and running to compare directly to Windsor.

AMD hasn't been talking a lot about mobile lately but the last word was that mobile and server would hit at the same time.

I am actually of the opinion that mobile should hit first since it is said that mobile revenue will overtake desktop this year.

no it won't

Though it is a good OC for this chip.

2.0GHz x 1.8 = 3.6GHz which is not easy for K8L chips

 

[BaronMatrix]

For those who were wondering the folks at AMDZone managed to get their 1.9GHz 3600+ to run stably at 3.1GHz with 1.48V and the retail HSF.

Linkage!

Looking at this result makes me think that Barcelona will do 80% OC. I mean Brisbane is just a shrink where Kuma will be from the ground up 65nm with all of that DSL stuff showing off it's ability.

We'll see though. Since AMD did mention 1.8x faster per dual core, they obviously have the mobile version of Barcelona up and running to compare directly to Windsor.

AMD hasn't been talking a lot about mobile lately but the last word was that mobile and server would hit at the same time.

I am actually of the opinion that mobile should hit first since it is said that mobile revenue will overtake desktop this year.

no it won't

Though it is a good OC for this chip.

2.0GHz x 1.8 = 3.6GHz which is not easy for K8L chips

So first everyone says how can AMD give numbers when it doesn't exist, but now everyone can say with certainty that it can't OC better than Brisbane?

 

[ajfink]

That's a pretty damn good OC. Like...better than 90nm Opteron good.

 

[BaronMatrix]

That's a pretty damn good OC. Like...better than 90nm Opteron good.

Definitely. 1.2GHz is HUGE. And they said it stays around 53C. Kuma will do at least as well probably better.

 

[qcmadness]

2.0GHz x 1.8 = 3.6GHz which is not easy for K8L chips

So first everyone says how can AMD give numbers when it doesn't exist, but now everyone can say with certainty that it can't OC better than Brisbane?[/quote]

So I would not say "2.0GHz x 1.8 = 3.6GHz which is impossible for K8L chips"

 

[BaronMatrix]

2.0GHz x 1.8 = 3.6GHz which is not easy for K8L chips

So first everyone says how can AMD give numbers when it doesn't exist, but now everyone can say with certainty that it can't OC better than Brisbane?

So I would not say "2.0GHz x 1.8 = 3.6GHz which is impossible for K8L chips" [/quote]

My point was that Brisbane is a shrink. AMD will probably do their usual and go for power conservation over clockspeed but from the papers on 65nm and Kuma being native 65nm, I would think a 15-20% increase in headroom would not be out of the question.

 

[qcmadness]

So I would not say "2.0GHz x 1.8 = 3.6GHz which is impossible for K8L chips"

My point was that Brisbane is a shrink. AMD will probably do their usual and go for power conservation over clockspeed but from the papers on 65nm and Kuma being native 65nm, I would think a 15-20% increase in headroom would not be out of the question.

Okay, I am a little bit convinced now

 

[papi4baby]

Where are the benchmarks?????? Maybe is not that stable as they say. Plus review about a AMD chip from AMDZONE is like a Intel review from Intel themselfs. Thanks Baron, it does look interesting but i'll wait for the benchmarks.

 

[BaronMatrix]

For quad core, yeah impossible on 65 nm --- for dual core, outside -- skin of their teeth in my opinion.... based on the current revision of 65 nm, it would be impossible.... time will tell how much they can sqeeze from 65 nm.

This is why the high-k announcement is so important.. AMD has the same gate oxide thickness as Intel -- this is why you see 3.0 GHz 6000+ running at TDPs of 120 watts and also why you see (at load) 65 nm AMD CPUs consuming about the same amount of power as Intel's 65 nm CPUs, though AMD uses fewer transistors.

AMDs load power should be a little higher because the IMC takes power. Again, we need to see a native part before making conclusions. Kuma is said to go to 2.9GHz at 89W/65W. Adding L3 and extra FP/SSE4, etc units means the process is rather good natively to remain in the same envelope.

I'm sure the 65W chips will be excellent OCers but I am only speculating based on what I've heard and seen.

 

[ajfink]

Like I've said, AMD's first 65nm parts are pretty rough, they're not even including some of the technologies they plan on implementing in "final" 65nm run. It's kind of like the warm-up lap. That fact that someone got a 3600+ to 3.1Ghz impresses the hell out of me.

 

[qcmadness]

AMDs load power should be a little higher because the IMC takes power. Again, we need to see a native part before making conclusions. Kuma is said to go to 2.9GHz at 89W/65W. Adding L3 and extra FP/SSE4, etc units means the process is rather good natively to remain in the same envelope.

I'm sure the 65W chips will be excellent OCers but I am only speculating based on what I've heard and seen.

The lower TDP ones are the worse OCers than the higher TDP ones.

 

[gOJDO]

For those who were wondering the folks at AMDZone managed to get their 1.9GHz 3600+ to run stably at 3.1GHz with 1.48V and the retail HSF.

Guys from AMDZone just provided a screenshoot of CPUz without a CPUz validation! Claims from a HARDCORE BIASED site like AMDZone are credible as much as Shakira's blog.
So, here I am, pushing their Brisbane to 5.95GHz on air:

:lol:
You know BaronBS, you are full of BS as always! Even if the CPUz screenshot is not fake, running a CPUz does not mean that the CPU is stable. We can say that it runs stably after it will pass 24 hours running two threads of Prime95 CPU torture, combined with RAM stability test.

I wonder what is the ambient temperature, because the system is not in ATX case and the CPU is surrounded with huge fans.

I'll skip commenting about the rest of your BS. Your mega fanboyism, blind optimisam and faithfull expectations about your beloved brand will make you to suffer more.

 

[halbhh]

your post was ok, but the last paragraph seemed like poltergeist.

 

[gOJDO]

Hahaha! :lol:
Their OC-ed Brisbane CPUz picture is PHOTOSHOPED! :tongue:

http://www.amdzone.com/pics/cpus/athlon64/65nm/x23600+brisbane/3600+oc.jpg
Download the image from their site and open it as a ASCII/text doucment. You'll reveal the secret behind their OC:
"Adobe Photoshop Elements 4.0"

 

[BaronMatrix]

Hahaha! :lol:
Their OC-ed Brisbane CPUz picture is PHOTOSHOPED! :tongue:

http://www.amdzone.com/pics/cpus/athlon64/65nm/x23600+brisbane/3600+oc.jpg
Download the image from their site and open it as a ASCII/text doucment. You'll reveal the secret behind their OC:
"Adobe Photoshop Elements 4.0"

Then I guess we can all ridicule them as frauds.

 

[BaronMatrix]

For those who were wondering the folks at AMDZone managed to get their 1.9GHz 3600+ to run stably at 3.1GHz with 1.48V and the retail HSF.

Guys from AMDZone just provided a screenshoot of CPUz without a CPUz validation! Claims from a HARDCORE BIASED site like AMDZone are credible as much as Shakira's blog.
So, here I am, pushing their Brisbane to 5.95GHz on air:

:lol:
You know BaronBS, you are full of BS as always! Even if the CPUz screenshot is not fake, running a CPUz does not mean that the CPU is stable. We can say that it runs stably after it will pass 24 hours running two threads of Prime95 CPU torture, combined with RAM stability test.

I wonder what is the ambient temperature, because the system is not in ATX case and the CPU is surrounded with huge fans.

I'll skip commenting about the rest of your BS. Your mega fanboyism, blind optimisam and faithfull expectations about your beloved brand will make you to suffer more.


You're STILL the little kid in your avatar. Your credibility with me DISAPPEARED FOREVER when you used my thumbnail as your avatar you sick stalker.

It will be BS when Kuma doesn't have 15-20%(200MHz/GHz) more headroom - that's just 3.6GHz from 3.0GHz. If AMDZone decided to cheat like that they should go down in flames.

 

[pete4r]

2.0GHz x 1.8 = 3.6GHz which is not easy for K8L chips

So first everyone says how can AMD give numbers when it doesn't exist, but now everyone can say with certainty that it can't OC better than Brisbane?

So I would not say "2.0GHz x 1.8 = 3.6GHz which is impossible for K8L chips"

I say you are a brainless idiot !!!

what part of 1.8 times of the PERFORMANCE do you not understand ?!?! your simple school boy theory is based on CLOCK SPEED and we here do not talk about CLOCK SPEED anymore. Just look at the Intel Core 2 Quad vs Core 2 Duo, in some Benchmark it is 2 times the performance than Core 2 Duo but I dont see no Core 2 Quad hits 4.0GHz or higher.

Quote me wrong if you got a source AMD said K8L will be 1.8 times the CLOCK SPEED of K8...

stupid idiots frustrates you and they think they are know all.

 

[shinigamiX]

2.0GHz x 1.8 = 3.6GHz which is not easy for K8L chips

So first everyone says how can AMD give numbers when it doesn't exist, but now everyone can say with certainty that it can't OC better than Brisbane?

So I would not say "2.0GHz x 1.8 = 3.6GHz which is impossible for K8L chips"

I say you are a brainless idiot !!!

what part of 1.8 times of the PERFORMANCE do you not understand ?!?! your simple school boy theory is based on CLOCK SPEED and we here do not talk about CLOCK SPEED anymore. Just look at the Intel Core 2 Quad vs Core 2 Duo, in some Benchmark it is 2 times the performance than Core 2 Duo but I dont see no Core 2 Quad hits 4.0GHz or higher.

Quote me wrong if you got a source AMD said K8L will be 1.8 times the CLOCK SPEED of K8...

stupid idiots frustrates you and they think they are know all.
They were talking about overclocking headroom.

 

[gOJDO]

You're STILL the little kid in your avatar.

Yes, I want to be a kid. Is there any problem?

Your credibility with me DISAPPEARED FOREVER when you used my thumbnail as your avatar you sick stalker.

I really don't care if you are going to respect me or not. Because you cry, I'll refresh your memory: you used to insult me, my nationality and my country, before I started falming you.

It will be BS when Kuma doesn't have 15-20%(200MHz/GHz) more headroom - that's just 3.6GHz from 3.0GHz.

You can't stop BS-ing!

If AMDZone decided to cheat like that they should go down in flames.

No! It will destroy Shakira's, Shitcentia's, 9nm's, LameNoobMike's and your life.

 

[dasickninja]

Its this another flame thread? I can be bothered to scroll beyond Barons reply.

 

[BaronMatrix]

Can't we all just get along? :lol: Actually I have to give props to Baron for accepting the possibility that the results could have been faked like he did and not cling on to the article like it was gospel. Just this once I say we give him a little bit of a break and cut him a tiny bit of slack.

Flame On. My day is not defined by your response to me or acceptance of me.

 

[BaronMatrix]

You're STILL the little kid in your avatar.

Yes, I want to be a kid. Is there any problem?

Your credibility with me DISAPPEARED FOREVER when you used my thumbnail as your avatar you sick stalker.

I really don't care if you are going to respect me or not. Because you cry, I'll refresh your memory: you used to insult me, my nationality and my country, before I started falming you.

It will be BS when Kuma doesn't have 15-20%(200MHz/GHz) more headroom - that's just 3.6GHz from 3.0GHz.

You can't stop BS-ing!

If AMDZone decided to cheat like that they should go down in flames.

No! It will destroy Shakira's, Shitcentia's, 9nm's, LameNoobMike's and your life.


I said sarcastically that I was surprised you had electricity in your cave in Macedonia. How is that an attack on anything but you, gDODO?

 

[MU_Engineer]

Three things:

1. Does Windows do a per-window screenshot of is it simply a full-screen dump? If it is the latter, one would have to crop the rest out to get just the CPU-Z window. That would require something like The GIMP or Photoshop.

2. 3.1 GHz on the chip sounds believable. CPU-Z verified shots from other sites have seen the Brisbane 4000+ do about 2900-3000 MHz. AMDzone only got 3-6% better, so it's entirely possible that their chip in fact did do 3.1 GHz. It was a very recently-produced chip and as we all know, AMD and Intel tweak with their processes over the runs. Also, perhaps AMDzone just got a chip that was like one of the CCBWE Opteron 165s and was from a batch that was more OC-friendly than other batches. Now if they'd gotten something like 3.3 GHz or better, I'd be suspicious, but 3.1 seems very likely.

3. I do frequent AMDzone occasionally and knowing them, I don't think they would lie. They are certainly a fan site, but if you sift through the forums and reviews, you'll see they try to get the facts right (even if they might - okay, do- spin them.) If the chip didn't overclock to 3.1 and hit the same 2.9 or 3.0 or whatever the other guys got, they would have published it anyway and talked more about how it would be a good, cheap overclocking CPU.

 

[epsilon84]

Hahaha! :lol:
Their OC-ed Brisbane CPUz picture is PHOTOSHOPED! :tongue:

http://www.amdzone.com/pics/cpus/athlon64/65nm/x23600+brisbane/3600+oc.jpg
Download the image from their site and open it as a ASCII/text doucment. You'll reveal the secret behind their OC:
"Adobe Photoshop Elements 4.0"

You're fcuking shitting me!

Are you certain they photoshopped it? That's a pretty serious accusation, and if true, this has immense implications on what (little) credibility they have left as a (fanboy) site.