AMD FX-8150 Bulldozer finally tested

[Monstru]

These days everybody is going back and forth regarding the new AMD architecture, and we have seen many architecture analysis, but now so many results. Well, here you go fellows, we put AMD FX-8150 head to head with 2600K in 8 benchmarks and 3 games. Enjoy!


AMD FX-8150 Bulldozer Preview - LAB501

[M.Beier]

THANK YOU BUDDY!
Finally some trustworthy results.

EDIT:
Zombie, for less skilled readers, it might be worth adding on the graphs "LOWER IS BETTER" "HIGHER IS BETTER" in Romanian obviously :-)

EDIT2;
The consumption graph seems wrong, are you sure you didnt mix up the results??
125W TDP on FX, but less than 100W actual?
95W on 2600K, but uses 130W?

[xBanzai89]

Wow, disappointed. Glad I returned my 990X mobo during the summer and bought a 2500K last week.

[gallag]

I didn't think it would be that bad,,,, bad drivers, bios or something?

[Sparky]

NDA is lifted now? I thought it wasn't for a few more days yet.

Anyway... something just seems so wrong with the whole thing. There are things that it does seem that the previous stuff (Deneb/Thuban) is faster in some respects. That's just messed up. Why would they do that

Maybe the refresh that is supposed to be here 1H 2012 will fix some stuff. Maybe. I don't know anymore what to think

I guess I may continue to rock my 955BE for a little while longer, put it back under water, and see if I can push the clocks a little higher. It may be better this way however, since I realized my bank account is a fair bit thinner than I thought it was.

[Monstru]

I didn't think it would be that bad,,,, bad drivers, bios or something?

Latest BIOS, latest drivers. Unfortunately this is how it performs... And that part in me that owned Athlon64 and was looking up at the FX series all my life is now crying

[Movieman]

NDA is lifted now? I thought it wasn't for a few more days yet.

Anyway... something just seems so wrong with the whole thing. There are things that it does seem that the previous stuff (Deneb/Thuban) is faster in some respects. That's just messed up. Why would they do that

Maybe the refresh that is supposed to be here 1H 2012 will fix some stuff. Maybe. I don't know anymore what to think

I guess I may continue to rock my 955BE for a little while longer, put it back under water, and see if I can push the clocks a little higher. It may be better this way however, since I realized my bank account is a fair bit thinner than I thought it was.

Your correct, NDA isn't lifted yet.

[cesariuth]

Monstru:

turbo core works fine?

[geo]

so time to run Anti Trust and breakup Intel then :/

[Kobaltrock]

Thanks for the results.

How many days until we see pile driver ?

Latest BIOS, latest drivers. Unfortunately this is how it performs... And that part in me that owned Athlon64 and was looking up at the FX series all my life is now crying

Any chance you can run some more tests from http://openbenchmarking.org/ ?

[M.Beier]

Oh snap... I forgot most important; CONGRATULATIONS with MOA WW!

[BeepBeep2]

THANK YOU BUDDY!
Finally some trustworthy results.

EDIT:
Zombie, for less skilled readers, it might be worth adding on the graphs "LOWER IS BETTER" "HIGHER IS BETTER" in Romanian obviously :-)

EDIT2;
The consumption graph seems wrong, are you sure you didnt mix up the results??
125W TDP on FX, but less than 100W actual?
95W on 2600K, but uses 130W?

The chip runs hot. Is there a problem with this?

The thermal design power (TDP), sometimes called thermal design point, refers to the maximum amount of power the cooling system in a computer is required to dissipate.

Yes.

There is a huge problem with this because I bet if you clocked up BD to where it consumed as much as the 2600K you would be more power efficient at same performance.

AMD designed a great CPU here, however it seems

A. It runs way too hot.
B. It doesn't clock high enough.

This is not a new Prescott, not netburst. It's not like this CPU consumes 150w vs the comparable i7.

I must say that I think GloFo's ty 32nm process is at fault here.

[RedBull78]

Is this a joke? No, seriously? How can it perform that low in RE game? 8 cores at over 4ghz and still loses to a quad core? Something's very fishy here..sorry.

[M.Beier]

The chip runs hot. Is there a problem with this?

The thermal design power (TDP), sometimes called thermal design point, refers to the maximum amount of power the cooling system in a computer is required to dissipate.

Look at the graph again.... It appears as Intel i7-2600K is using 129,72W, while FX uses 93,12W... They are mixed up, the i7-2600K uses the 93,12W If you translate the text below, it is pretty clear that the graph has been mixed up.

[Frontl1ne]

Your correct, NDA isn't lifted yet.

I'm surprised more people haven't picked up on this yet. Why does Lab501 get to break NDA?!

[BeepBeep2]

Look at the graph again.... It appears as Intel i7-2600K is using 129,72W, while FX uses 93,12W... They are mixed up, the i7-2600K uses the 93,12W If you translate the text below, it is pretty clear that the graph has been mixed up.

Alright

I'm really disappointed now...

It runs hot, clocks poorly (worse than SB with worse IPC than Deneb), and it seems an X6 @ 4.3 will be able to beat it in...uhh, everything.

[PerryR]

It runs hot, clocks poorly (worse than SB with worse IPC than Deneb), and it seems an X6 @ 4.3 will be able to beat it in...uhh, everything.

And none of that makes any sense...........What the hell is going on here?

[Kobaltrock]

I'm surprised more people haven't picked up on this yet. Why does Lab501 get to break NDA?!

Someone set their server clock ahead a few days.
Either that, or they were never under NDA, and they got the CPU from someone that sold it a bit too early.

[Mechromancer]

brokenNDA.jpg

Your new avatar Monstru!

Those scores look ridiculously low. Isn't the PII X4 980 and/or 1100T faster in most of those cases?

[PerryR]

brokenNDA.jpg

Your new avatar Monstru!

Those scores look ridiculously low. Isn't the PII X4 980 and/or 1100T faster in most of those cases?

It doesn't add up. Based on this review, this processor is worse than the previous generation.

[BeepBeep2]

brokenNDA.jpg

Your new avatar Monstru!

Those scores look ridiculously low. Isn't the PII X4 980 and/or 1100T faster in most of those cases?

In single thread yeah, 980 will win hands down. Actually, your 940 might be competition for this...
In multithread (and single thread), yeah, 1100T has this thing pretty beat.

Considering I'm buying one of these regardless of air/water performance, which is lower than the 4.8-5.0 stable I expected for air and 5.2 I expected for water...(I got called a pessimist for estimating this) within launch week I hope AMD pulls another card out of its sleeve at the last minute.

[Peakr]

Handbrake is a terrible video encoding benchmark. Any chance you could try x264 bench?

[M.Beier]

What is going on here is Intel is a MUCH bigger company, has ALOT more researchers, and AMD was close to facing bankroupcy, ofc. the investers want value for their money... The segment is of less importance....

AMD is performing very well in laptops....
AMD is performing very well in HTPC...
AMD is performing very well in netbooks...
AMD is BLOWING for desktop/workstation...
I cannot comment server segment, as I do not have suffecient knowledge.

Small Form Factor, I'd pick AMD for the time being, I wouldnt even have any doubt that Zacate OWNS ATOM... The Sandy Bridge is really strong - Ivy Bridge is not much stronger, just lower power consumption at same performance... AMD is behind on this segment, and as Monstru clearly expressed, it truely hurts remembering the greatness of AMD64 FX (FX51,53,55,57,60) and then seeing this, mocking the legacy bigtime.

[gallag]

If this is true, not that I am doubting you monstru just finding it hard to believe, but if it's true then this is serious for AMD.

[Dumo]

Too bad