AMD FX-8150 Bulldozer finally tested

[Brice MJ]

Yeah, right...

[PerryR]

Yeah, fastest back in 2008.

[Brice MJ]

Yeah, fastest back in 2008.

We had Nehalem back in 2008.

[Cecil]

The problem is, they would have been much more competitive with Sandy Bridge if they had shrunk X6 (I'm not sure why guys keep saying "add two cores!"), added 4-6MB between L2 and L3 caches, and sped stock speeds up to 3.6-4.0 single thread. I'm fairly certain that is doable within a 95-125w TDP on 32nm, and they should have been able to figure out some way to gain clockspeed...intel has had no problems with it, and AMD did it well from transition to 65nm to 45.

X6's that would OC on water 4.6-4.7 Ghz on 32nm +3% IPC boost from extra cache or adopted Llano's STARS core + L3, would have gained ~5-10% IPC over current gen and would have been a lot better than this.

Thats definitly what they should have done, to get them selves in a better situation for right now. They could have saved BD a little longer to work on it more.

But, its AMD, and they didnt. They want to try to change the game too much when they need to focus on playing it better.

[PerryR]

We had Nehalem back in 2008.

This beats that......Barely. At least the original models, and non-enhanced editions.

[Cecil]

This beats that......Barely. At least the original models, and non-enhanced editions.

Uh, 920 is faster then Deneb/Thuban, this is slower then Deneb and Thuban.

[Brice MJ]

This beats that......Barely. At least the original models, and non-enhanced editions.

FX8150 can barely come close to Phenom II X6 1100T in low-threaded apps, so no, comparing it to 2008's Core i7 965 3.2GHz is not even funny.

[radaja]

Thats definitly what they should have done, to get them selves in a better situation for right now. They could have saved BD a little longer to work on it more.

But, its AMD, and they didnt. They want to try to change the game too much when they need to focus on playing it better.

yep,i think they just got sideswiped doing the new uarch and unmature node shrink simultaneously(very risky).and now we are seeing the salvaged product(the best AMD could make of the situation)
hopefully we will see Big improvements upcoming with this new uarch

[Hornet331]

Seems another guy snaged a 8120:
http://forums.anandtech.com/showpost...&postcount=192

Is currently putting the rig together.

[BeepBeep2]

Uh, 920 is faster then Deneb/Thuban, this is slower then Deneb and Thuban.

And an X6 @ 3.6 with 6-7% IPC increase average from extra L2,L3 + Llano's "STARS" core in 95/125ww TDP envelope would beat Nelahem in single thread and smash it in multithread workloads...

I'm pretty sure that would have made a lot of us more than happy enough.

[PerryR]

Uh, 920 is faster then Deneb/Thuban, this is slower then Deneb and Thuban.

I thought 1090t was essentially even with 920?

http://www.anandtech.com/bench/Product/146?vs=47

Also, I thought the 8150 was marginally faster than the 1090t?

FX8150 can barely come close to Phenom II X6 1100T in low-threaded apps, so no, comparing it to 2008's Core i7 965 3.2GHz is not even funny.

A 965 is an extreme edition processor, which I said it doesn't beat. I think you mean the 940?

[Oliverda]

Anyway for AMD it's clear now that they should have done a new socket from 0.
May be to fit and work better with BD arhitecture.

[AliG]

And an X6 @ 3.6 with 6-7% IPC increase average from extra L2,L3 + Llano's "STARS" core in 95/125ww TDP envelope would beat Nelahem in single thread and smash it in multithread workloads...

I'm pretty sure that would have made a lot of us more than happy enough.

See that's what makes me question these results a bit. If anything, isn't BD an augmented version of Llano? So shouldn't at the very least the multithreaded performance be drastically better due to CMT?

[ice_chill]

On the other hand, lets be greatfull we still have someone who competes with Intel.

[AliG]

On the other hand, lets be greatfull we still have someone who competes with Intel.

Perhaps, but I personally find that to be a losing attitude. As the consumer we should never settle for what the industry hands us, that's the whole point of supply and demand

[PerryR]

Where's JF?

[xBanzai89]

Where's JF?

Likely responding to all the emails he is getting about this or went on a vacation where there is no internet till the 12th.

[radaja]

Where's JF?

this is what he said yesterday at ocn

A. I could never comment on that.

B. ESPECIALLY during quiet period.

so im sure he wont comment on this due to quiet period

[Iconyu]

Where's JF?

If these numbers are true, I think he's getting drunk.

[celerity]

If Lab501 is right then this is fake:

http://www.ozeros.com/2011/10/nuevos...e-el-i7-2600k/

I don't think AMD present fake numbers to customers, they may select apps that runs well on the product that is presented but they can't fake it

Hold on... there is something really weird with the graphs:

(1) http://img.ozeros.com/noticias/2011/...500K_2600K.jpg

(2) http://img.ozeros.com/noticias/2011/...150_slide3.jpg

In the first Bulldozer is ~20% faster than 2600K in wPrime 32M. However, in the second it's 20% slower than 2600K. Same for other graphs, e.g. in the first 2600K is faster @ Winrar 4, second shows Bulldozer being faster. How come?

[PerryR]

I just realized that slide from 2007 said fastest cpu core in "history!" I thought it said mainstream cpu core.

LOL. What was AMD smoking back then?

[Cecil]

I thought 1090t was essentially even with 920?

http://www.anandtech.com/bench/Product/146?vs=47

Also, I thought the 8150 was marginally faster than the 1090t?




A 965 is an extreme edition processor, which I said it doesn't beat. I think you mean the 940?

A 920 is faster then a Thuban in ST apps, Thuban catches up in really MT stuff. But thats at stock clocks with the 920 at 2.6 and the Thuabn at 3.2. With both at 4Ghz the 920 is much faster. Also, most 920s will continue on to clock higher.

[AffenJack]

Where's JF?

Look at AliG's Signature and at the Benches so far, looks like he might have a lot to do tellin customers where BDs IPC increase went.

Was the donaminhaber video test already here?
http://www.donanimhaber.com/islemci/...o-inceleme.htm

[PerryR]

A 920 is faster then a Thuban in ST apps, Thuban catches up in really MT stuff. But thats at stock clocks with the 920 at 2.6 and the Thuabn at 3.2. With both at 4Ghz the 920 is much faster. Also, most 920s will continue on to clock higher.

I was mostly referencing to stock clocks, I already knew that the 920 will win in most apps when over clocked.

[Hornet331]

Look at AliG's Signature and at the Benches so far, looks like he might have a lot to do tellin customers where BDs IPC increase went.

Was the donaminhaber video test already here?
http://www.donanimhaber.com/islemci/...o-inceleme.htm

Haven't seen it... but the graphs in the vid basically show the same thing again... 8150 slower then X6 or barly beating it... (gona make a few screen caps)