AMD or Intel?

[Martin Cracauer]

For integer workloads Intel's Core2-base Xeons outperforms K8 (the
old-school AMD64) by about 25-30% per clock per core.  K10 seems to be
5-15% faster than K8 for integer workloads (I hope to run my benchmark
suite on one thi week or weekend).

However, tasks that use multiple cores and have threads on cores
communicate a lot see both AMD architectures close the gap.  

Paul Pathiakis wrote on Mon, Sep 10, 2007 at 10:17:40AM -0400: 
> Be very, very careful in purchasing Core 2 Duo.  There are major 
> problems with the chip that have been documented across the board.  

These have been blown out of proportion by Theo.  Can you point to a
demonstratable case with current Linux or BSD kernels?

I also highly doubt that the AMD design is much more bug-free.
Although it might.  Core2 has very complicated caches, K10 is still
simpler.

On the other hand, if you want K8 or K10 in a modern SMP mainboard you
have to live with NVidia for chipsets, and the socket F boards all
have the MPC55 SATA controller, which iirc is unsupported by both BSD
and Linux.  MPC65 moved to AHCI so all is well - but there are no
socket F boards with that SATA controller.

One real advantage of Socket F over Socket 771 is that you can use
normal registered DDR2, which goes dirt cheap both on ebay and new.
Socket F sees mandatory FB-DIMMs, which are expensive and hot, and
boards with more than 8 DIMM slots are rare.  For socket F 2x8 DIMM
slot boards and plenty and cheaper.

Martin
-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Martin Cracauer <cracauer at cons.org>   http://www.cons.org/cracauer/
FreeBSD - where you want to go, today.      http://www.freebsd.org/