Any chance of more threading in the future

[Amak]

Sounds right. Wait for Bulldozer, it is supposed to compete with the i7 head to head.

[Squall Leonhart]

no it isn't.

[abdo123]

Sounds right. Wait for Bulldozer, it is supposed to compete with the i7 head to head.

so not going to happen

[pseudonym]

We don't really benefit much from SSE 4, but every little helps. And we don't rely on heavy floating point by choice, the VUs just are floating point units.

[DaMan]

If AMD can get BD running at the ~3.5GHZ-4GHZ they want it to than it should be competive against a stock clocked SNB. Will probably still get owned clock for clock though.

[Squall Leonhart]

its still not competive against a stock SB.

[rama]

Yea Squall, let's at least wait for the benchmarks and comparisons on the final silicon, okay?

[Squall Leonhart]

already seen benchmarks.
all i can say is for apps that can scale threads easily its decent, but for apps that have limited multithreading (like most) its not worth looking at.

[rama]

You've seen leaked, faked, server software benchmarks.

[Squall Leonhart]

No, the benchmarks are legitimate B2 results from reviewer samples.

you'll see when the NDA expires anyway, the chips are just really crappy for thread limited applications.

the results from the ES on chiphell reflect the performance from my Reviewer contacts results. (And chiphell has rarely been wrong about chip performance)

chiphell has it at 3.2Ghz and gets a score of 23s in superpi
the czech guy who had it at 4+Ghz got 17-19s

a Nahalem Core i7 920 gets 11s at 4Ghz
a 2600k gets <10s at 4Ghz