• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

FEATURED Bulldozer rumors

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Status
Not open for further replies.
Throwing some rough numbers into this if you look at the folding forums, they are saying SB is about 33% faster than current i7's. So if BD is 50% faster then we are looking at them being 20% faster than Sandy :clap:

But yes just all rumors.....
 
While it's definitely not something to be positive about, it makes sense. Eight Amd cores ~ 150% of four intel cores. Technically, that's still not winning clock/clock. It mentioned being roughly 20% faster than the 980x which means two extra cores is still only 20% faster. It's much better than PII, but we'll have to see real results.
 
FAH points is a poor benchmark, the points per day goes up wildly as cpu speed improves, it's far from linear.

SB is ~10-15% more efficient clock/clock compared to nehalem. (Which may well make 33% more ppd!)

If AMD is doing what I think they're doing with bulldozer, it should be very good.
 
While it's definitely not something to be positive about, it makes sense. Eight Amd cores ~ 150% of four intel cores. Technically, that's still not winning clock/clock. It mentioned being roughly 20% faster than the 980x which means two extra cores is still only 20% faster. It's much better than PII, but we'll have to see real results.

Are there any other dreams of mine you would like to destroy?
 
Bahaha, trust me. I'm definitely one of the people who want Amd to take the performance crown back, but we'll have to see definitive results to be sure. ;)
 
While it's definitely not something to be positive about, it makes sense. Eight Amd cores ~ 150% of four intel cores. Technically, that's still not winning clock/clock. It mentioned being roughly 20% faster than the 980x which means two extra cores is still only 20% faster. It's much better than PII, but we'll have to see real results.

But still,if you get really technical, two of AMD's Bulldozer Module cores only make up one processing core.
 
Last edited:
I don't think it works like that... (unfortunately.)

According to the link, one module is two cores (really, two half cores sharing cache).

Bulldozer will be available in three variants - four, six and eight core (or two, three and four module).
 
Dolk if you understand all of that then wow your smart! lol it is interesting but way over my head. lol but i get the basics of it.
 
I am still waiting for 0 wait state computing. When that happens even the slowest will have a fast feel.

Parallel processing is a natural evolution but is only a step in the journey.
 
Not to burst any bubbles here but Parallel computing has been around for a long time, in fact any computer with a multi core processor or multiple processors, or a graphics card with multiple shaders routinely does parallel processing (any time larger task is split into multiple smaller ones and executed at the same time). If Bulldozer's new instruction set and accompanying compiler can get closer to automatic parallelization (converting sequential code into multi-threaded or vectorized or both) then AMD could have a real winner on their hands. Basically an SMP on each module ;).

Also check out Gustafson's law which supersedes Amdahl's law.
 
Last edited:
after reading the Bulldozer 20 questions again I saw this
We are working with all of the key compiler vendors to help ensure support of Bulldozer. We are spending a lot of time working with the Open64 compiler folks to make sure that there is support, as well as the PGI Group, GCC compiler and of course, Microsoft®.

AVX will require applications to be recompiled in order to take advantage of 256-bit floating point (either ours or our competitors).

I can’t comment on the ICC compiler, I recommend asking them that question.

My bad they don't actually have a compiler:
We (AMD) work with all of the major compiler vendors to integrate support in. We also have the AMD Core Math Library that can be used when you compile applications, but it is not a compiler.
 
Last edited:
Here's a very good read for those inclined, AMD's Bulldozer Microarchitecture.

While most of that article went right over my head, it was a great read. Probably the biggest thing to stick out to me were these:

AMD first managed to break into the server market in 2003 with the K8, thanks to the 64-bit extended instruction set and system architecture choices such as an integrated memory controllers and on-die interconnects. Intel certainly helped out as well, since the Pentium 4 was a distinctly unfriendly product for servers. AMD eventually hit an impressive 25% market share, and over 50% share in the lucrative 4-socket server market by 2005/6. The key to AMD's success was that they provided exactly what consumers wanted (x86-64 and good server performance), while Intel was distracted with Itanium and the P4. In essence, they found an area where Intel could not (or would not) focus, and then put all their efforts into addressing customer needs and were able to change the rules of the game.
(Page 2, Paragraph 1)

AMD's failure here was largely a result of trying to match Intel's superior manufacturing and resources head-on.
(Page 2, Paragraph 2)

Philosophically, Bulldozer seems to learn from the lessons of the past decade. AMD is stepping back from the pursuit of single threaded performance, to emphasize throughput.
(Page 2, Paragraph 4)

In essence, Amd has figured out that they cannot go head to head with intel's design ideas and win. As much as I don't like to admit it, it's true. Single threaded performance is just stellar on intel. Amd used to be winning in the server market because they were focusing on something intel wasn't. They [Amd] lost a majority of that market by trying to compete head to head with intel's design ideas.

With Bulldozer, it seems they have put a lot of work into circumventing the pitfalls of the past and have moved onto a much more efficient, overall, architecture. :thup:
 
Status
Not open for further replies.
Back