Hyperthreading test on sandy bridge: makes a huge difference

[jebo_4jc]

I had a chance to compare speeds on the same WU, on the same mobo, with identical CPUs, other than one was an i5-2400 which has no HT, and a Xeon E3 1220 which has HT.

On a 6900 WU:

E3 1220 (with HT):
TPF: 35:36
PPD: 28,296

i5 2400 (no HT):
TPF: 48:30
PPD: 19,221

Big difference

 

[pjkenned]

The retail Xeon E3-1220 does not have hyper-threading BTW. Probably bad for Google if people see this post about the engineering sample chip you are using.

Xeon E3-1220 Benchmarks
Xeon E3-1230 Benchmarks

100 MHz clock speed difference between the two which is <3% and does not account for the differences.

For the full lineup information I have started an Intel Xeon E3-1200 series information thread.

I am actually doing a few folding benchmarks of the chips + a X8DTH-6F with dual E5606's, Core i7 Sandy Bridge CPUs, and the new E3-1280 (benchmarks like the above for the E3-1280 should go up Monday).

I will say that the Core i7-2600K @ 4.4GHz is significantly faster then the dual E5606's.