AMD to launch desktop Bulldozer "Zambezi" at E3 Show in L.A. (June 7-9)

Quote:

---

Originally Posted by ubuntu83

Everyone with internet access can send him a scan of a 500 Euro note. :p:

---

Post system still exists :welcome:

Quote:

---

Originally Posted by XRL8

Um, 4ghz chip is always almost better than 2x2ghz ones.
And yes, it can be more than 2x as fast even.Depends on the work hes doing.FSB speeds, IMC speed,memory management, threading etc etc.
All this talk however is too mystical.40% more raw horsepower may or may not get us 40% higher bench score.That depends.
However people should understand that statement is most probably some huge simplification, as on say "on average".Im pretty sure BD will get often MORE than 40% ,and often much less.
You could say that 2500k has 40% more performance than a core2quad.But again, that depends on the work cpu is doing.
Simply taking into account that it has 2 more cores, gives you around 30% more computational power.Most probably higher speed, better turbo etc.And better IPC.And new instructions.
AMD would have to actually TRY to NOT get a sizably faster processor with all these changes.
Lets remember that Phenom II is OLD.They had lotsa time architecture wise.

---

Of course we are talking with all other components the same. If you only change the CPU speed 100%, you won't get 100% higher scores in benches.

Youre oversimplyfying everything ;-).Youre talking about theoretical thoroughput not "performance".
And while yes its very rare, you can have higher than clockspeed increase boost in performance.
Computer system is very complicated software/hardware "organism" ,with the amount of complexity it becomes harder and harder to predict.
Example:
http://www.legionhardware.com/images...70/L4D2_02.png
http://www.legionhardware.com/images...70/L4D2_02.png
Look at the 2.2 to 2.4ghz increase, actual "bench" score goes higher than the increase in clocks.Most probably at 2.2 cpu was running to slow and was choking rest of the system.
In some crossfire situations you got more than 100% scaling also.It happens.
That is OT though, "performance" is a real life metric, its not thoroughput.So if AMD States 40% higher performance, you can be sure it will get in atleast one scenario 40% higher real life SCORE than phenom II did get.

Quote:

---

Originally Posted by XRL8

Youre oversimplyfying everything ;-).Youre talking about theoretical thoroughput not "performance".
And while yes its very rare, you can have higher than clockspeed increase boost in performance.
Computer system is very complicated software/hardware "organism" ,with the amount of complexity it becomes harder and harder to predict.
Example:
http://www.legionhardware.com/images...70/L4D2_02.png
http://www.legionhardware.com/images...70/L4D2_02.png
Look at the 2.2 to 2.4ghz increase, actual "bench" score goes higher than the increase in clocks.Most probably at 2.2 cpu was running to slow and was choking rest of the system.
In some crossfire situations you got more than 100% scaling also.It happens.
That is OT though, "performance" is a real life metric, its not thoroughput.So if AMD States 40% higher performance, you can be sure it will get in atleast one scenario 40% higher real life SCORE than phenom II did get.

---

Is it oversimplyfying when I say that overall benches won't reflect the pure performance increase linearly? You provided a benchmark that proves my point exactly! One Core i7 @4GHz is twice as fast as one i7 @2GHz. But your bench don't show that at all. It would be bad logic to say that the i7 2GHz only needs to be 15% faster to compete with the i7 4GHz based on your bench.

In the exact same way it's bad logic to say that BD only needs to be 35% faster than Phenom II X6 to compete with 980x. I say it needs to be alot faster than that.

Of course we won't have total linear scaling with frequency, but the closest we can get to linear scaling to make this example, there will be small flucturations as your bench shows. I'd say that your analysis that the 2.2GHz Phenom was to slow is wrong since we would see that affecting the 2GHz part as well. I say it probably has more to do with asynchronous frequencies with the NB and L3.

And of course there is a few scenarios where the bench is able to show the real difference, in one of these scenarios an BD might be 50% faster than Phenom II X6, but don't you think that 980x is alot faster than that in those scenarios as well? Benches showing 980x more than 50% faster than X6 is all around.

Closing this thread for now as we haven't had any "news" in it for 2 weeks. For general discussion, please visit the AMD section and start a thread on speculative pricing.