Donanimhaber has pulled out this article :shrug:
Printable View
Let me put it this way. There's no way those benchmarks are right. I'm not familiar with all of them, but I know for sure that POV Ray and X264 Pass 2 scale extremely well with more cores. In very well threaded applications, Thuban will just about match Intel's quad cores. In those two particular benchmarks, they're showing Zambezi performing slightly worse than Sandy Bridge. That's an octa core running slower than a quad core with almost half the die size. All I want out of Zambezi is to perform slightly better than Gulftown.
Where did they say that? They said both Sandy Bridge-e and Bulldozer should be out by thanks giving so a comparision whould be possible then. He did not say that Bulldozer wouldn't be released even earlier than that. If Bulldozer is released October 12th and SB-e at thanksgiving then his statement is correct.
No one has been able to point out where I am wrong. If you want to call me a troll then you should be able to point out what's wrong with my post.
X264 Bulldozer Development http://forum.doom9.org/showthread.ph...46#post1528746
Which "tests" are you talking about, and by Nehalem, are you referring to the 980x? A processor which even gives the 2600k a run for its money in some tests (not overall)?Quote:
And that's against mid-end CPUs in handpicked tests. How can you call that a success? As I said, it gets beaten by Nehalem which is three years old. It's simply a flawed architecture.
Always keep an eye on motherboards manufacturers. :)
http://www.biostar.cn/app/en-us/t-se...s.php?S_ID=554
更新CPU代码,支持AM3+ B2-F / B2-G 步进CPU
B2-F ????:confused: B2-G ????:confused:
I doubt BD will perform better than gulftown... that would be expecting to much. Even SB-E has a hard time passing gulftown zith double digits.
IIt would be more realistic if they can compete with gulftown in lowthreaded while having a small advantage to the 2600 in highly threaded.
I would guess FMA will be much more important for X264 and any other codecs.
Obviously B2F wasn't good enough for Zambezi, so that is why it was delayed. But since Interlagos was needed by close partners at first, they could ship some bad chips to Cray and friends and call it a server launch. Even though nobody else is actually getting them. And we know two types of Zambezi have been leaked, the one they'll ship first, and a new one in Q1 2012. I'm guessing B2G is the first one they'll ship, and a B3 is coming which will replace it ASAP. The preceding text is complete and utter speculation!
it could be that server chips were doing just fine without the respin since they are using no where near the same clocks as consumer chips
here, Opteron146 posted a schedule :D
is it around the corner? man, BD was supposed to be q2-3 of this year, and we are yet to see it released. So if piledriver(Vishera) is q3 next year, I don't want to laugh about it, but .... around the corner indeed :)
MicroCenter employee is stating to me they will be able to sell BD on the 12th of October.
The main problem is the 32 nm process tech which is not matured enough and there is a capacity problem too. I think both of these issues will be solved in few months.
It's always difficult to start a brand new microarchitecture on a brand new and complex (SOI + gate first HKMG) process tech.
Attachment 120555
:)
Would be awesome, finally a launch coming?