AMD showcases World's fastest CPU

I think the thread title is a bit misleading, we don't know if AMD has the world's most fastest CPU, but I do congratulate them for making a CPU that can overclock greatly. Now if they just released this already instead of teasing us with all these information I'll be happy.
 
John Vuong said:
I think the thread title is a bit misleading, we don't know if AMD has the world's most fastest CPU, but I do congratulate them for making a CPU that can overclock greatly. Now if they just released this already instead of teasing us with all these information I'll be happy.
Click to expand...

kyles video plus a bunch of others are proof, they broke the 8Ghz barrier on LN2.. now think what they might be able to do with liquid helium.. given that they were able to pull an extra 800mhz out of a phenom II x6 between LN2 and liquid helium... sandy bridge isn't a threat to break the record any time soon either since they hit a wall at ~6Ghz and won't run with liquid helium.
 
sirmonkey1985 said:
kyles video plus a bunch of others are proof, they broke the 8Ghz barrier on LN2.. now think what they might be able to do with liquid helium.. given that they were able to pull an extra 800mhz out of a phenom II x6 between LN2 and liquid helium... sandy bridge isn't a threat to break the record any time soon either since they hit a wall at ~6Ghz and won't run with liquid helium.
Click to expand...

well, to be fair, you can still tweak the bclock a lither after reaching max multi... (at which point, it will make a big deal, lol).
 
well it looks like the chip's got legs.
can't wait to pop one in my CHV.

one thing i was curious about, is it 'normal' to disable most of the cores for these OC runs? cpuz shows 2 cores in the clip.
i understand of course running 8 cores at that speed would be incredibly stressful for the cpu. not that it matters i suppose as all the other cpu's in top 10 were dual core celerons.

one thing did piss me off tho, there was 50 or so chips on the test bench. that's 50 chips the AMD consumer will never see...:mad:
 
AMD in the past has stated that IPC was improved over the K10. K10's world record is 7.1Ghz, BD reached 8.4Ghz on an engineering sample and on just a few days of tinkering.

There is no doubt in my mind a $266 will get you one sweet CPU. Now AMD hurry up and take my money.

I am predicting I'll be running a 6Ghz BD pretty soon here.
 
they got 4.8g on water, and 5.8 on phase change. anything more was ln2 and finally liquid helium and that was done with 6 of the 8 cores disabled on nothing but suicide runs. the water results are good, they are about on par or even better than a good 2600k. how the two stack up will be the key as it may take more clock speed to equal the intel chips.
 
images


With those prices. Get me better than a 2600K and I'll dump the rig
in the sig.
 
ZacharyS said:
Because as has been known for a very long time, simply having more geegahurtz doesn't always mean it has better performance.
Click to expand...

You're the only person who's talking about performance here. I was talking about frequencies and Bulldozer being a better overclocked than Sandy Bridge. I never mentioned performance. Performance is an unknown at this point so why even debate it?

SB can't even beat 6Ghz, here we have Bulldozer doing 8.4Ghz on engineering samples. That's quite a bit more headroom.

I was responding to a person who said BD's overclocking is on par with SB, when in reality it looks like BD will hit on water cooling what SB hits on LN2.
 
Last edited:
im going by the video that was taken at the event. in the video they only achieved 4.8 under water.
 
mrbigshot said:
im going by the video that was taken at the event. in the video they only achieved 4.8 under water.
Click to expand...


We achieved clock frequencies well above 5GHz using only air or sub-$100 water cooling solutions.
Click to expand...

http://www.pcper.com/news/Processors/AMD-Bulldozer-Processor-hits-8429-GHz-New-World-Record

ENGINEERING SAMPLE

Also to author of the article goes on to say:
I was at the demo, and wrote this article. The 5GHz air results were actually with all cores enabled and a workload applied - not just a CPU-Z verification. I would have put that in the article, but I hadn't thought of anyone raising that question. Thanks.
Click to expand...
 
Last edited:
If I have to read one more time "available for sale shortly" I'm going to blow a gasket. Even buxom blondes seducing octogenarian billionaires don't string them along this long!
 
corrupted223 said:
http://www.1up.com/news/pentium-4-cpu-overclocked-8ghz

So these kind of tests dont count then?
Click to expand...

What do you mean? the link you have shows 8.0 ghz where as the this recent article shows 8.4ghz

Personally I think this is good, obviously their IPC is not going to be better than intel but if they can OC better than intel then maybe people like us can find more value in the AMD CPUs just like the good old athlon days.
 
W.Feather said:
for what? a system built by a sub-par company? ... cheaper to buy the same exact parts and assemble it your self(will last longer as well...)
Click to expand...

You aren't excited about them unveiling a powerful PC with Bulldozer, and possibly hinting at an upcoming release? I don't plan on buying a IBP PC any time soon. LOL.
 
sirmo said:
I was responding to a person who said BD's overclocking is on par with SB, when in reality it looks like BD will hit on water cooling what SB hits on LN2.
Click to expand...

Yeah, and use twice as much power, and be just as fast. Clock speed is irrelevant when you're discussing completely different cores, especially with one with a massive advantage in IPC. 5GHz Sandy might be as fast as 6-7GHz Bulldozer.
 
iconiK said:
Yeah, and use twice as much power, and be just as fast. Clock speed is irrelevant when you're discussing completely different cores, especially with one with a massive advantage in IPC. 5GHz Sandy might be as fast as 6-7GHz Bulldozer.
Click to expand...

Why would it use twice as much power? This isn't P4 we're talking about here. This is BD. AMD knows how to make a higher clocked solution that uses less power (Evergreen vs Fermi).

As far as performance goes I don't trust any of the leaks so far. Every single one of them looks shady. So really the performance may be awesome or just average, no-one knows. But even if it was on par with Phenom II. With this sort of OC-ing potential and 8 cores, I am pretty happy.

If it can reach the IPC of Nehailem... gg Intel.
 
Last edited:
mrbigshot said:
they got 4.8g on water, and 5.8 on phase change. anything more was ln2 and finally liquid helium and that was done with 6 of the 8 cores disabled on nothing but suicide runs. the water results are good, they are about on par or even better than a good 2600k. how the two stack up will be the key as it may take more clock speed to equal the intel chips.
Click to expand...

My 2600k hits 5.1GhZ with a Corsair H50. 4.8GhZ is achievable with enthusiast-class air cooling on the 2600k, there's a forum with a 5GhZ SB club where a lot of guys are using Noctua and Thermaltake coolers.

8.4GhZ is awesome, though. Kinda wishing I waited for BD to arrive, but when June came and went the upgrade itch got to me. I'd really like to take a shot at OCing one of these.
 
LAiN_OfTheWired said:
My 2600k hits 5.1GhZ with a Corsair H50. 4.8GhZ is achievable with enthusiast-class air cooling on the 2600k, there's a forum with a 5GhZ SB club where a lot of guys are using Noctua and Thermaltake coolers.

8.4GhZ is awesome, though. Kinda wishing I waited for BD to arrive, but when June came and went the upgrade itch got to me. I'd really like to take a shot at OCing one of these.
Click to expand...

Mrbigshot's post is misleading. He watched 2 minutes of a video and grabbed random data and stated it as the max they were able to achieve. In reality BD engineering samples reached well above 5Ghz on just air. Come release, I think 5.5-6Ghz will be what BD gets on air.
 
Even if it only has a lifespan measured in hours on 2 hands, that's still pretty cool.

It's kind of disappointing that a record allows 3/4 of the cores to be disabled though. Of course scrap from poor yield might make that configuration a real market possibility. :p
 
pxc said:
Even if it only has a lifespan measured in hours on 2 hands, that's still pretty cool.

It's kind of disappointing that a record allows 3/4 of the cores to be disabled though. Of course scrap from poor yield might make that configuration a real market possibility. :p
Click to expand...

Previous record holder Intel Celeron 352 had all the other cores disabled also.

http://valid.canardpc.com/records.php
 
sirmo said:
Mrbigshot's post is misleading. He watched 2 minutes of a video and grabbed random data and stated it as the max they were able to achieve. In reality BD engineering samples reached well above 5Ghz on just air. Come release, I think 5.5-6Ghz will be what BD gets on air.
Click to expand...

6Ghz is pushing it a wee bit.. i'm figuring air cooling may get 5.2-5.4Ghz but i don't see 6 Ghz on air. they were only able to hit 5.8Ghz on phase change just sitting in windows showing a cpu-z screen. my prediction is 5.2Ghz will be the norm for high end air cooling, 5.5-5.6Ghz on high end water cooling..

pxc said:
Even if it only has a lifespan measured in hours on 2 hands, that's still pretty cool.

It's kind of disappointing that a record allows 3/4 of the cores to be disabled though. Of course scrap from poor yield might make that configuration a real market possibility. :p
Click to expand...

the basic rules for record overclocking is that you have to be able to get into windows and take a screenshot of cpu-z. if you can't do that then the overclock fails. theres no rules on how many cores have to be active though.
 
sirmonkey1985 said:
6Ghz is pushing it a wee bit.. i'm figuring air cooling may get 5.2-5.4Ghz but i don't see 6 Ghz on air. they were only able to hit 5.8Ghz on phase change just sitting in windows showing a cpu-z screen. my prediction is 5.2Ghz will be the norm for high end air cooling, 5.5-5.6Ghz on high end water cooling..
Click to expand...

Engineering sample. Good batch Sandy Bridge can reach 5.2Ghz on air. BD seems to OC much better due to it's longer pipeline.
 
sirmo said:
Engineering sample. Good batch Sandy Bridge can reach 5.2Ghz on air. BD seems to OC much better due to it's longer pipeline.
Click to expand...

these were all cherry picked Engie samples though so you still have to take that into account.
 
sirmonkey1985 said:
these were all cherry picked Engie samples though so you still have to take that into account.
Click to expand...

All the previous world record holders were mature production steppings or mature fab process, in case of these BD samples neither is true, also AMD claims they were not hand picked overclocks but random CPU engineering samples, something to consider.
 
That's all fine and dandy but most of us, if not all, don't have the resources to get anywhere near that. Don't get me wrong it's a great record but it's just publicity for AMD.
 
sirmo said:
All the previous world record holders were mature production steppings or mature fab process, in case of these BD samples neither is true, also AMD claims they were not hand picked overclocks but random CPU engineering samples, something to consider.
Click to expand...

they were hand picked out of an allotment of like 70 processors.. they admitted it in the video kyle posted. look as much as you i'd like to see 6 ghz possible on air.. but its not going to happen. maybe with the updated stepping used with the 8170 coming out next year it may be possible but it ain't happening with these.
 
random CPU engineering samples
Click to expand...

To me that is a scary thought. Does that mean they still did not have any production silicon at the time of the demo? Or does that mean the production silicon is not as good as the engineering samples as regards to overclocking?
 
Last edited:
drescherjm said:
To me that is a scary thought. Does that mean they still did not have any production silicon at the time of the demo? Or does that mean the production silicon is not as good as the engineering samples as regards to overclocking?
Click to expand...
They could be like the Phenom II TWKR 42, engineering samples with a very high leakage current, whereas the production silicon has much lower leakage.
 
You must log in or register to reply here.
Back
Top