Actually, as far as the video compression benchmark goes, I strongly
suspect that's the new SMT implementation making itself known, as
video decode/encode doesn't tend to be memory latency bound, and the
thing only had one memory channel working properly so it wasn't
bandwidth making the difference.
And, well, most people wouldn't max out a 2Ghz A64 either, but we've
been able to say that for probably the last decade about the current
state of hardware... I'd like to see someone playback 1080P video on a
P3-500 though :p
True enough DDR3 pricing is going to make it a very high end option,
since the first ones to market will say Xeon on them though... and
will be DDR3 rather than FB-Dimm DDR2.. sign me the hell up for a Q408
Mac Pro with 8 cores plz :)
-JB
On 6 Jun 2008, at 14:27, <[EMAIL PROTECTED]> <[EMAIL PROTECTED]
> wrote:
I wasn't wowed =S
Yes, its a decent performance boost but then that was always
expected once they got the memory controller into the die....
The issue here is......most people aren't maxing out their current
Core2 CPUs, let alone need the power of these.
Plus DDR3 is still 2-4 times more expensive then DDR2 which is going
to make uptake of these quite slow to begin with I expect.
Not knocking Intel, just do not see the Core2 sales dropping by much
based off of these results.
Regards
Jason
