This thread discusses the Content article: Intel Core i7: Dual-Channel vs. Triple-Channel Memory Mode

I don't believe youz!

I puts my fingers in my ears and shakes my head whilst going NO NO NO!

How about asking Intel, which makes an X58 board and the Core i7, to explain this?

How about you explaining how the tests ensured you were testing a three-way interleaving vs. two. I've never heard of some of the programs. But then I'm mainly a Unix person.

I know what WinRar is; isn't it I/O bound? You would have to use files loaded into memory for that to test whether it's taking advantage of full memory speed possibilities, wouldn't youz?

How about writing a tiny C program that malloc's a MB of core memory or whatever the kids are calling it these days...then do some read and read/write access loops against three consecutive fullwords and see if THAT demonstrates a difference. Is the interleave three 32-bit fullwords or 64 on this 64-bit processor?

It is not clear to me that you have diagnosed memory itself.

--
http://grandscheme.org

Ha, you are wrong.

Using a SiSoftware Sandra benchmark to measure memory bandwidth preformance:

http://www.techspot.com/review/124-intel-core-i7-920-940-965/page5.html

QUOTE:
Clearly moving to an on-die memory controller has paid dividends as the Core i7 920 squeezed out a memory bandwidth of 17GB/s in SiSoftware Sandra using nothing more than DDR3 memory running at 1066MHz. Running three of these modules in triple-channel mode has huge performance benefits, as we have just witnessed.

Overclocking the Core i7 965 EE along with the DDR3 memory allowed for a memory bandwidth of 21GB/s, which is very impressive. When compared to the Core 2 Quad Q9650 running dual-channel DDR3-1333 memory, the Core i7 920 delivered well over twice the memory bandwidth.

--
http://grandscheme.org