AMD Bulldozer Engineering Sample CPU Overclocked to 4.63GHz

[bridito]

Are they talking about Trinity?

Trinity is not mentioned anywhere in the article and this is the only mention of BD... So your guess is as good as mine.

 

[Accord99]

The good news is, if they keep changing the stepping we might eventually get to a b-17 and can call it the "superfortress" (and yes, I know that I just mixed my bomber metaphors)

Technically, AMD would need to get to the 29th revision to get a "superfortress". At 17, they're just at the flying fortress stage.

 

[podspi]

Yea I thought the B1 and B2 were scrapped and they are working on b3 right now.

Last I heard, B2 was the magic number, although I think server is going to get B1. That implies there is a clocking problem D: but no IPC problem at this point

WTF? I just saw this and it's dated today:

"AMD's next-gen architecture, code-named Bulldozer, is due next year."

http://seekingalpha.com/article/276697-amd-s-fortunes-rise-with-llano-but-risk-remains

It's a fairly legit financial site... are they wrong or...

(Gawd, I hope not...)

I saw that too, pretty sure they're referring to Trinity. Server should be out in Q3 of this year.

 

[PlasmaBomb]

that must not be the same type of run. the BD screenshot says 1m29s. those are 32M runs and not 1M run like the BD.

You need to look more closely at the screenshot, because it doesn't say that, it has been edited.

The time is for pi to 1M and the time taken is 1X.Y29 seconds (where 0<X<9 and 0<Y<9), meaning it is between 10.029 and 19.929 seconds... which is a massive range, and unless it's closer to 10 isn't brilliant for 4.6GHz...

The voltage required also looks like 1.55V (again it has been edited).

PS. as others have stated, it's a fake...

 

[bridito]

Last I heard, B2 was the magic number, although I think server is going to get I saw that too, pretty sure they're referring to Trinity. Server should be out in Q3 of this year.

It would seem rather silly for a financial analyst to mention every other new AMD CPU family and just skip forward to Trinity while calling it BD. If that's the case, the dude should hit the unemployment lines.

So either he's a total dolt, or he knows something we don't.

Disclosure: NOT trying to put out the fire with gasoline. ()

 

[cantholdanymore]

It would seem rather silly for a financial analyst to mention every other new AMD CPU family and just skip forward to Trinity while calling it BD. If that's the case, the dude should hit the unemployment lines.

So either he's a total dolt, or he knows something we don't.

Disclosure: NOT trying to put out the fire with gasoline. ()

Trinity will have BD core

 

[formulav8]

Well 2nd 1/4 is almost over and no cpu. They definitely ran into trouble somewhere...

 

[bridito]

Trinity will have BD core

Yes, but the BD itself is due either late Aug. or early Sept. as of the last AMD announcement.

Well 2nd 1/4 is almost over and no cpu. They definitely ran into trouble somewhere...

Let's pray they didn't run into another 6-9 months of trouble!

 

[GaiaHunter]

Again, 16th of July [H] lan party. Kyle is convinced BD will be there.

 

[podspi]

It would seem rather silly for a financial analyst to mention every other new AMD CPU family and just skip forward to Trinity while calling it BD. If that's the case, the dude should hit the unemployment lines.

So either he's a total dolt, or he knows something we don't.

Disclosure: NOT trying to put out the fire with gasoline. ()

People make mistakes. This financial analyst either:

1) Just made a mistake
or
2) Just wrote something he shouldn't have (and shouldn't have known either)


As of right now, the last information AMD has made public is that server Bulldozer is on track for Q3. That might change (god forbid, haven't I lost enough money already?! ) but AMD hasn't said anything to that effect, so I'm voting 1) unless that guy has insider info (in which case he just did a bad thing).


Also, isn't Trinity BD Enhanced, not Bulldozer? If Trinity was going to be the first BD-based product, that would imply Bulldozer V1.0 got cancelled D:

BTW, I do agree with the sentiment of the article. Llano is looking to be a nice product, but CPU performance is very lacking. I think this is going to keep it out of high-end mobile gaming machines. Not that that actually matters this year, but Trinity can't come soon enough. Intel isn't going to sit still, and for people who aren't gaming the i3 looks to me to be a better mobile CPU.

 

[bridito]

People make mistakes. This financial analyst either:

1) Just made a mistake
or
2) Just wrote something he shouldn't have (and shouldn't have known either)


As of right now, the last information AMD has made public is that server Bulldozer is on track for Q3. That might change (god forbid, haven't I lost enough money already?! ) but AMD hasn't said anything to that effect, so I'm voting 1) unless that guy has insider info (in which case he just did a bad thing).


Also, isn't Trinity BD Enhanced, not Bulldozer? If Trinity was going to be the first BD-based product, that would imply Bulldozer V1.0 got cancelled D:

BTW, I do agree with the sentiment of the article. Llano is looking to be a nice product, but CPU performance is very lacking. I think this is going to keep it out of high-end mobile gaming machines. Not that that actually matters this year, but Trinity can't come soon enough. Intel isn't going to sit still, and for people who aren't gaming the i3 looks to me to be a better mobile CPU.

Yeah, I'm leaning towards the DUH! school of journalism as well. The guy was talking out of his nether regions. I don't "think" that it was insider info that he inadvertently spilled but it really goes to the issue of: what happened to fact checking before running a story?

All we can do is keep praying to the AMD gods that they can finally get a really strong performing BD out. I'm really looking forward to the first independent test results that haven't issued from The Photoshop Benchmarking Utility!

 

[ElFenix]

eh? seeking alpha is basically huffpo, they set up a blog site and generate revenue from user content. they're not necessarily authoritative or knowledgeable. that contributor has only been making posts there for the last month. google's first page doesn't dig up anything on the guy.

post a comment asking if the guy is really referring to trinity or to at least cite his source for BD being pushed to next year

 

[996GT2]

2500K@ 5.0 GHz. Add about 0.5 seconds for 4.63 GHz.

 

[bridito]

post a comment asking if the guy is really referring to trinity or to at least cite his source for BD being pushed to next year

Done. Got "Your comment is being reviewed by our moderator and will be posted shortly. Thank you for your patience." Let's see if he replies... REALLY? I DIDN'T KNOW THAT?

 

[bridito]

2500K@ 5.0 GHz. Add about 0.5 seconds for 4.63 GHz.

So even if the first blank is a 0, that BD ES is about 25% slower than a 2500K? For AMD's sake it better be wrong.

 

[Phynaz]

So even if the first blank is a 0, that BD ES is about 25% slower than a 2500K? For AMD's sake it better be wrong.

What makes you think it would be any faster single threaded?

 

[bridito]

What makes you think it would be any faster single threaded?

Beats me. Some of the wild rumors running around had it trouncing the 2600K by far. But the bottom line is that there are no accurate performance figures on BD anywhere so far. At least as far as I've seen.

 

[soulsavior]

So even if the first blank is a 0, that BD ES is about 25% slower than a 2500K? For AMD's sake it better be wrong.

Super Pi isn't really useful for comparing processors anymore, at least across architectures. The instruction set it uses has been obsolete since like the Pentium 2. I'm no computer engineer, but I'm fairly certain no CPU really optimizes for the instruction set used with Super Pi anymore, so while better scores could translate into real world applications, they also may not.

It's kinda crappy people are apparently addicted to releasing leaks in Super Pi form. I'd assume it's because it's easy to fake with its more or less plain text output.

 

[gevorg]

1.29sec makes it very easy to conclude that its fake.

 

[grimpr]

SuperPi is absolute crap compared to wprime and y-cruncher, which will soon come out optimized for AVX & Bulldozers FMA unit.

 

[Spikesoldier]

super pi is fake

1.29s for 1m is bull

super pi is a bad indicator of stability.

 

[dac7nco]

"It's a FAAAKE!"

-Senator Vreenak

Glory to the Dominion!

 

[podspi]

Beats me. Some of the wild rumors running around had it trouncing the 2600K by far. But the bottom line is that there are no accurate performance figures on BD anywhere so far. At least as far as I've seen.

Don't forget those four extra cores...

Bulldozer doesn't need to be faster in ST performance than Intel. Bulldozer just needs to not get embarrassed in ST tasks like K10 does.

 

[Idontcare]

Super Pi isn't really useful for comparing processors anymore, at least across architectures. The instruction set it uses has been obsolete since like the Pentium 2. I'm no computer engineer, but I'm fairly certain no CPU really optimizes for the instruction set used with Super Pi anymore, so while better scores could translate into real world applications, they also may not.

It's kinda crappy people are apparently addicted to releasing leaks in Super Pi form. I'd assume it's because it's easy to fake with its more or less plain text output.

Its easy to fake a screenie of SPi, this is true, but its popularity really stems from the fact that since it is a single-threaded app it can be used on core-disabled rigs for suicide benches.

Its not very strenuous, passing SPi is one step above being able to idle without blue-screening. So you can get it to bench on remarkably unstable hardware.

 

[SonicIce]

"It's a FAAAKE!"

-Senator Vreenak

It's REAL!