• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD FX Series and A Series First Performance Projections Surface

Doesn't show clock speeds either :S

I was hoping for a lot better than that. I will wait for official benches
 
Just like most "leaked" slides in recent memory, I'll just wait until release date to see what performance is like.

I really can't see us having another Phenom (1) on our hands.
 
Well if its as fast as a 2600k, thats a pretty damn good chip, though might depend on the size.

I've been saying for a while, we're not just looking at cores anymore, you have to consider whats in them. 2600k = 4 x 4 issue cores = 16 instruction maximum per clock.

Bulldozer, 8 x 2 issue cores = 16 instruction maximum per clock.

Keep in mind that filling a 4 issue core is MUCH harder than filling a 2 issue core, so getting 16 instructions is easier from 8 2 issue cores, the problem then becomes, needing 8 threads. Of course plenty of other stuff is constantly done in the background while loading the other cores, less sharing and less interuptions are a good thing for performance. HT helps the 2600k act more like a Bulldozer, as you can push 8 2 issue threads through a 2600K.

But thats VERY basic, each issue can be faster than on another core. if you've got 8 threads that can fill two issues, it will likely run significantly faster on the Bulldozer than the 2600k, though certainly not always.

It depends how Vantage works really as too if it favours one architecture over another. Also its a system benchmark not just a cpu benchmark meaning limits occur elsewhere, both cpu's in those benchmarks(bulldozer and 2600k) could have hit other hardware limits so show the same score.

Slightly more relevant is the Bulldozer vs hex core with the same gpu benchmark, a significant increase in a not altogether cpu limited benchmark is really quite impressive.

Maybe most importantly is cost, AMD tends to focus all their marketing over the overall feel of your computer at any given cost, rather than flat out performance. Are they hinting that a hexcore Phenom + 6670 will cost close to a 2600k, which will be in the same price range as a Bulldozer + 6670?


Ultimately AMD and Intel are at architecture's now that are as different as AMD and Nvidia. If its an 8 core, or a 88 core, those are just terms, they don't signify performance. Bulldozer matches or beats the 2600k, and costs the same or less and it is a win for AMD. If its slower than a 2600k but costs the same or more to produce AND has to be sold for less due to performance, it will be a massive fail for AMD.

I think in the same vain as Nvidia vs AMD now on the GPU front, the different architectures will shine in different software, so one benchmark won't come close to telling the whole story.

The most likely situation is, worst case situations bulldozer is very close toa 2600k, and best case(8 thread, not good in HT) Bulldozer could have a significant lead, if they can provide all that inside a slightly better priced chip than Intel, its a win for us. Cheaper mobo's, cheaper chip, same or more performance, whats to hate.

Something I find quite, irksome about Sandybridge is the IGP, paying for it even if you don't, or even CAN'T use it seems like a huge huge waste.
 
Last edited:
Interesting (and good for AMD) if it ends up close to the 2600k. But then Sandy Bridge-E will more than likely destroy it.

All the same, can't wait for some proper reviews & benchmarks.
 
A Core I7 2600K beats a Core i7 990X in PCMark Vantage:

http://www.tomshardware.co.uk/core-i7-990x-extreme-edition-gulftown,review-32126-4.html

This indicates the benchmark does not make very good use of more cores. Hence,the top end Bulldozer CPU is doing a good job if it is matching a Core i7 2600K.

Edit!!

Even when you consider the 100MHZ increase in clockspeed and an extra 4 threads the Core i7 2600K is under 10% faster than a Core i5 2500K in PCMark Vantage:

http://www.neoseeker.com/Articles/Hardware/Reviews/Intel_i7_2600K_i5_2500K/8.html
 
oh dear, what a let down, really hope they are not accurate/real

??

So it can match a Core i7 2600K in a benchmark that can only really use 4 cores effectively. Even a Core i5 2500K has similar scores to a Core i7 2600K in PCMark Vantage. A Core i7 990X scores less than a Core i7 2600K in the benchmark too.

If this benchmark is true then even a Core i7 2600K is going to be much slower in applications like video encoding and rendering.
 
Last edited:
As I thought, looks like bulldozer has the upper hand in multithreading. But I use loads of single threaded software unfortunately, so hopefully some more benchmarks will be leaked out soon. Really am tempted with the 2600K though :p.
 
It is worrying that they need to use AMD CPU + AMD GPU combo graphs compared to Intel CPU + IGP to look superior.

If an 8-core Bulldozer (possibly the fastest one) only matches an i7-2600K in multi-threaded applications (in which case it's almost certainly slower per thread), it had better overclock damned well and be a hell of a lot cheaper.
 
A Core I7 2600K beats a Core i7 990X in PCMark Vantage:

http://www.tomshardware.co.uk/core-i7-990x-extreme-edition-gulftown,review-32126-4.html

This indicates the benchmark does not make very good use of more cores. Hence,the top end Bulldozer CPU is doing a good job if it is matching a Core i7 2600K.

Edit!!

Even when you consider the 100MHZ increase in clockspeed and an extra 4 threads the Core i7 2600K is under 10% faster than a Core i5 2500K in PCMark Vantage:

http://www.neoseeker.com/Articles/Hardware/Reviews/Intel_i7_2600K_i5_2500K/8.html

It is generally held that hyperthreading adds around 10% when compared to a single thread. This is apparent in the benchmark you have cited, and is not as you are suggesting down to the nature of PC Mark etc, more the nature of hyperthreading itself
 
It is generally held that hyperthreading adds around 10% when compared to a single thread. This is apparent in the benchmark you have cited, and is not as you are suggesting down to the nature of PC Mark etc, more the nature of hyperthreading itself

Look at the first link. The 3.46GHZ six core Core i7 990X has around a 20% higher PCMark Vantage score than the quad core 3.2GHZ Core i7 960. So basically for around an 8% increase in clockspeed the two additional cores are giving around an additional 12% improvement in performance.

In the second link the 3.5GHZ Phenom II X4 970 beats the 3.2GHZ Phenom II X6 1090T. The Phenom II X4 is not only clocked higher but each core has higher IPC due to the increased L3 cache per core over the Phenom II X6.

Anyway,since Bulldozer is being released next month hopefully we will soon know how well it performs.
 
Last edited:
Even if bulldozer matches or abit faster than a SB. Its AMD's first new architecture since Athlon days, AMD can tweak, improve and mature its new architecture.
 
Back
Top Bottom