Quad Core benchmarks
 

There is also no benchmark for Core 2 Quad, which I imagine could be great for GIMPS.

[QUOTE=jinydu;93870]There is also no benchmark for Core 2 Quad, which I imagine could be great for GIMPS.[/QUOTE]
There may come some change to that situation in the next two weeks ;-)
The quad core turns ought to be cheaper than other solutions if one calculates the ratio of cpu power per price unit (only one motherboard, disk, motherboard...)
More important : the ratio of cpu power per energy cost is also higher.

[url]http://www.mersenne.org/bench.htm[/url]

Quad core is slower than dual core?! :w00t: :w00t: :w00t:

[QUOTE=jinydu;94612][url]http://www.mersenne.org/bench.htm[/url]

Quad core is slower than dual core?! :w00t: :w00t: :w00t:[/QUOTE]

You mean faster, don't you?

The QX6700 is essentially two E6770s stitched together, and if you look at the iteration times for each on the benchmarks page, they are almost identical. The next system I build for GIMPS will be a quad core, but the price is going to have to come down considerably before I do so. Also, from other forums, it seems there is a substantial heat issue when all four cores are at 100%.

For all but one of the FFT sizes, the quad core is slower than the dual core. I would have expected the quad core to be much faster.

[QUOTE=jinydu;94650]For all but one of the FFT sizes, the quad core is slower than the dual core. I would have expected the quad core to be much faster.[/QUOTE]
jinydu, I think you are completely ignoring the clock speed and comparing a higher clocked Duo to a lower clocked Quad. If you compare like for like, you'll see that there is pretty much no difference (or even that Quad is faster).
:rolleyes:

EDIT:- Why would you think a Quad core will be "much faster"? These are single threaded benchmarks -- number of core shouldn't affect it one way or the other.