Linus Torvalds (torvalds@linux-foundation.org) on 2/5/09 wrote:
---------------------------
>The much more interesting issue is if they fixed anything
>else than the platform?
>
>SpecInt runs basically close-to-100% out of the cache, and
>of all the architectures out there, Itanium is way way
>behind the competition. Improving on the platform will do
>absolutely nothing to that, since it isn't even
>relevant.
>

At 1GHz, Alpha EV7 was about 10% faster than Alpha EV6 on SPEC INT2000. The EV7 is just the EV6 core with an all-new "platform."

If Tukwila debuts at 2.2GHz, then I'll estimate it will score around 25 on INT2006 (base), which is close to the scores for the 3.33GHz Core 2.

>Look at the SpecInt2006 numbers:
>- Core i7-965: 36
>- p570 (4.7Ghz): 21.7
>- Itanium (1.66GHz/18MB): 17.0
>
>Now, the Core i7 number is obviously much more recent than
>the other two, but hey, that's what Tukwila will be facing.
>(And quite frankly, you can find old Core 2 numbers that
>are just under 30, so even if you want to go back to the
>previous generation, you're really not changing the argument
>at all).

Keep in mind that libquantum is a broken benchmark. Core 2 Quad scores are inflated by 25% due to it.

>And if Tukwila only improves on the platform, not the
>core (and frequency), then it will be stuck at roughly
>the same performance. Don't tell me that the platform
>matters that much for SpecInt when you have a 18MB L3. I
>just don't believe you.

I'll put my chips on Tukwila being released at 2.0GHz, and achieving 2.4GHz in it's lifetime. How do you feel Tukwila will fare against Nehalem in the quad socket market on workloads like TPC-C/H, SAP-SD, SpecJBB, etc?