Replies to post #60808 on Intel Corp (INTC)

[alan81]

I thought Barcelona was supposed to beat the INTC products by 40% in specfp_rate at the same clock speeds...
oops.
--Alan

[chipguy]

Why is AMD so far behind Intel in single thread performance?

Good thing AMD doesn't sell processors for client computing.

Oh wait...

[wbmw]

Re: Wasn't Harpertown supposed to beat Barcelona in SpecFPrate ?

That would have been embarrassing for AMD, indeed, if they can't even with on the most bandwidth hungary benchmark they've touted since the beginning.

But as it is, I found a better submission than Chipguy, which is only behind by 4%. #msg-28185707

Actually, this is downright pathetic for AMD, that in the best case benchmark out there they are within a 5% margin. What a dog of a new architecture that in the best case possible, it only wins by 4%, while Xeon wins by over 30% in int_rate_base! Unbelievable....

Re: Beating the similarly clocked and priced Tigerton is another surprise in SpecINTrate.

Too bad that's not a fair comparison for Tigerton, since the E7330 has only 6MB of cache, while the E7340 also runs at 2.4GHz, and has 8M of cache.

http://www.intel.com/products/processor/xeon7000/specifications.htm?iid=products_xeon7000+tab_specs

The extra 2M is enough to make a tie in performance:

http://www.spec.org/cpu2006/results/res2007q4/cpu2006-20071113-02615.html

Yes, I know, that means Barcelona at 2.3GHz can perform as well as Intel's 2.4GHz, second bin of Tigerton in specInt_rate_base. What a huge win for AMD to be just about even with Intel, even in the 4-socket space, given a brand new architecture... Not.

[spaarky]

Wasn't Harpertown supposed to beat Barcelona in SpecFPrate?

It has been up until now. I congratulate AMD on at last producing the Barcelona they needed 3 quarters ago.

-=spaark=-