What is to be expected of amd bulldozer?

[postmortemfear]

just very curious if i should wait for 8 cores or just get six cores for now. what features with amd bulldozer have and anyone know the expected price?

 

[Lmeow]

Bulldozer around a half-year to a year away IIRC, no one knows the pricing, just that there will be hex core and octo core processors. If you need a computer now, I suggest you should just buy an AMD Phenom II X6/X4 or Core i7/i5, if you can wait/e.g. do not need one, then I'd wait for another year or so.

 

[loneninja]

Bulldozer is still close to a full year away, possibly longer. It'll defineatly out perform Phenom II, but by how much no one knows right now except AMD engineers. Pricing defineatly won't be known, as pricing will reflect upon performace/price figures of processors in a year, which no one knows that either.

Point is, unless something is right around the corner, it's not worth waiting for because there is always something else coming within another year.

 

[Dekasav]

Yeah, Bulldozer is expected (hopefully) to do very well, but actual pricing and performance numbers are a long way off. Bulldozer is a good year+ off.

 

[postmortemfear]

this what i found on wiki

Bulldozer
"Orochi" (32nm, 8-core)

* 4 AMD Bulldozer modules
* 8KB L1 data cache(8 way associate) + 4kuops trace cache/core, 64KB L1 instruction(2 way)/module
* 256KB L2 cache/module
* More than 8MB L3 cache share by all module and GPU.
* new dual channel multiple threading, can be up to 8 cores/24 threads
* DDR3
* AM3r2
* 10-100 watts[2][3]
* Tentative release date Q1/Q2 2011

"Zambezi" (32nm, 16-core)

* 8 AMD Bulldozer modules
* DDR3
* AM3r2


i probley will just end up waiting but hop these specs are true

i have q6600 and 3.4ghz and 2x 5770's in crossfire just really wanted faster for video encoding and a insane 3dmarks06 scores lol

 

[Raidur]

I have high hopes for bulldozer. I'm hoping they'll perform near SandyBridge, but Phenom II is performing right under Yorkfield (IPC). Hopefully matching/beating nehalem IPC.

If they don't intel is going to sell SandyBridge for $450-$2000. =-o

 

[Lmeow]

More like over $9000!

 

[FALC0N]

I expect significant IPC improvements on the cpu core. This is the first significant redesign of the AMD cpu core architecture in a while. And this is where they have been trailing Intel.

If AMD will catch or overtake Intel any time soon, thats when it will likely happen. The last time a major core revision was undertaken, it was intel in 2006 with the core 2 duo.

 

[amk09]

More like over $9000!

Ahh begone with yourself meme, you do not belong on this forum!

 

[hundredislandsboy]

Why wait for 8. If your work can make full use of 6 cores now, why wait?

 

[fazers_on_stun]

I expect significant IPC improvements on the cpu core. This is the first significant redesign of the AMD cpu core architecture in a while. And this is where they have been trailing Intel.

If AMD will catch or overtake Intel any time soon, thats when it will likely happen. The last time a major core revision was undertaken, it was intel in 2006 with the core 2 duo.

Hmm...

FALC0N wrote:

Core 2 didn't just magically appear in 2006. It was developed INCREMENTALLY as an extension of the P3 architecture and later Pentium M for laptops and was simply moved to the desktop.

It appears to be a huge jump ONLY because it replaced Netburst. But if you follow its parallel development history, it improved INCREMENTALLY.

Two-faced much there, Mr. "Calculus professor"??

Now I'm guessing you're a 2nd-year engineering student somewhere, since you seem to think Calculus is the peak of mathematical knowledge. Hate to break it to ya, but you still have differential equations to master, and if you're any good, operational mathematics where you study the design of transform kernels such as the Laplace transform as used in EE.

Anyway, back on topic, yes BD promises to improve integer performance significantly, which is where AMD fell behind Intel with the advent of C2D and the 4-issue core.

 

[FALC0N]

Hmm...

Two-faced much there, Mr. "Calculus professor"??

Read what you quoted one more time.

---------------------------------------------------
"Core 2 didn't just magically appear in 2006. It was developed INCREMENTALLY as an extension of the P3 architecture and later Pentium M for laptops and was simply moved to the desktop.

It appears to be a huge jump ONLY because it replaced Netburst. But if you follow its parallel development history, it improved INCREMENTALLY."
----------------------------------------------------

Like I say right there, the performance jump from netburst to core was real, but it was not a technical jump. That was all about poor intel decision making and not technical advances.

The quote itself answers your accusation.

Now I'm guessing you're a 2nd-year engineering student somewhere, since you seem to think Calculus is the peak of mathematical knowledge. Hate to break it to ya, but you still have differential equations to master, and if you're any good, operational mathematics where you study the design of transform kernels such as the Laplace transform as used in EE.

This is a clear example of your problem in almost every topic of discussion. Comprehension.

The only thing your paragraph and my remarks that you purport to be responding to have in common is the word Calculus. Its so far off topic, you could substitute a paragraph about the man on the moon or the tooth fairy and it would lose nothing. Might actually make more sense that way.

I even wrote out that my remark was analogous and you still bobble it.

Anyway, back on topic, yes BD promises to improve integer performance significantly, which is where AMD fell behind Intel with the advent of C2D and the 4-issue core.

You and your obsession with the "4 issue core".

Well, they have had 4 years to study the intel execution core. They could just copy it and see a 25% speed boost. So yes, it would be a huge disappointment at this juncture if they didn't at least match the current Intel execution core performance.