Core i7 and chess

[bob]

As hyper-threading technology allows the acceleration of multi-threaded applications which chess programs are multi-threaded?

First came Crafty. Then any engine that is "deep something" such as deep fritz or deep junior, etc. Also it doesn't require "threads" just "multiple processes". I have used both although I now use threads. Some use processes (ala fork() in unix) for design reasons...

[bob]

How then does non-chess software get a speedup on the Intel Core i7? I was reading of the claims of 1.7 to 3.5 speedup for commercial applications on the new RHEL 5.3 Presumably all those applications are not written specifically for the Intel Core i7 and have been out for years and years?

In internal testing, the Red Hat Engineering Performance Group has measured exceptional gains with the new Nehalem processors, with unaudited results showing gains of 1.7x for commercial applications and gains up to 3.5x for high-performance technical computing applications compared to the previous generation of Intel processors.

Better internal processor design features. Just compare core-1 to PIV to see a _huge_ speed increase while running at a _slower_ clock speed. Ditto for core-1 vs core-2. My core-2 2.0ghz is about 4-5x faster than my PIV xeon 2.8, just measuring one processor to one processor...

[terminator]

If i compiled crafty on the Core i7, would any compiler switches improve its performance? What would happen if Intel have all future processors hyper-threaded? Would chess continue to be non-multithreaded?

[bob]

If i compiled crafty on the Core i7, would any compiler switches improve its performance? What would happen if Intel have all future processors hyper-threaded? Would chess continue to be non-multithreaded?

Not quite sure what you mean. Crafty _is_ multi-threaded. But until you can take a single processor, and turn on hyper-threading, and see well over 30% NPS increases, hyper-threading will only hurt chess engines. A well-written program will not have many of the bottlenecks that hyperthreading helps, so the benefit of the extra logical processor (nps increase) is far outweighted by the additional search overhead an extra thread incurs. And you can't get rid of that extra search overhead...