Benchmarking for memory bandwidth?

[maxSe]

Okay... In anticipation of arrival of my BH-7, I've been running and recording some final benchmarks with my IT7 Max2... I'm not sure BH-7's going to help as I feel that I'm reaching the limits of my 2.4B SL6EF, but that's another story...

I like to know what would you guys recommend for benchmarking the memory bandwidth... I'm currently running it at 181fsb with 4:5 memory ratio. Only benches I've ran so far were 3D Mark 2001SE & 2003. I get very simiar scores running at 3.266GHz(181fsb/4:5mem) as running @3.429GHz(190fsb/1:1mem). Of course the score is slightly higher in 190fsb, but I'm not going to lose any sleep giving up few points in 3DMark. (complete scores will be posted sometime tonight )

I'll be mainly using the 2.4B/BH-7 as gaming rig (not a big FPS guy, but every now and then I enjoy some ), and occasional mpg/avi encoding. I just want to see what would be more benficial for me to run... Any thoughts?

 

[chizow]

Grab Sisoft Sandra for free in the thread stickied at the top of this page. Theoretical max test, but that's really a bandwidth test is good for, theoretical max.

Chiz

 

[THUGSROOK]

a real memory benchmark?
i dont think one truely exists.

id say memtest86~
time how long it takes to run 1 loop, all 11 tests.

that test would be 100% memory dependant. (i think)



<EDIT> sandra sux dont bother with that

 

[maxSe]

chizow, hmmm... I have a copy of Sandra already... I'll look for the Theoretical max test module tonight (didn't even know there was such...). I've only used Sandra for memory bandwidth scores...

THUGS, well I run the memtest, but it's mainly for stability... I'll give that a thought.. But I was hoping for some "real world" use for higher memory bandwidth... just trying to see if it's really that beneficial to use 4:5 or 3:4 ratio other than "because my memory kicks @ss"...

 

[mrman3k]

Those are the two benchies I use for my pathetic memory bandwidth.