new sandra beta

[acejj26]

hey guys, over at lost circuits, they did a review of the palomino vs the thunderbird core and in it they have sandra memory benchmarks. check these out!!! maybe they rewrote the memory bandwidth algorithms or something because that is more than a 100% increase over previous tests on their 8KHA review. Does anyone know how or why this happened?? According to the benchmark, it is faster than the dual channel Rambus solution in the i850 chipset. Your guys' thoughts??

 

[NOS440]

Well I would say its a software bug for sure just like the bug when the P4 came out and with SSE enabled it scored 50% more than it was physically possible.

 

[VBboy]

I don't know what to say, I just wanted to see three Duke icons in a row

 

[Rage187]

LOL

you freakin clones..

 

[formulav8]

hi

 

[NFS4]

<< Well I would say its a software bug for sure just like the bug when the P4 came out and with SSE enabled it scored 50% more than it was physically possible. >>


Why not read this article before jumping to conclusions about a software bug:

http://www.vanshardware.com/articles/2001/july/010720_Sandra_Bandwidth/010720_Sandra_Bandwidth.htm

 

[Slacker]

Who let rage and NSF4 in here?