Fwd: AMD x2 chips

[Bill Broadley]

On Wed, Feb 15, 2006 at 08:38:37AM -0800, Bryan J. Smith wrote:
> mailing-list-amd64 at smithstone.co.uk wrote:
> > I am considering getting a AMD x2 based machine and was
> > wondering if people could confirm that 
> > if i installed the 64bit of RHEL the system would be
> > recognized as a dual system and not just a single core
> > system
> 
> >From the standpoint of not only the APIC, but even the HyperTransport
> interconnect, you have _two_physical_ processors.  No "fancy

Not sure what the definition of physical is here.  It's different than
say a dual opteron single socket system.  There is no hypertransport
between the two cores.

> bridging" or anything is required for AMD Athlon64 x2 or Opteron xx5
> products, the actual interconnects are to _two_different_, _physical_
> ICs.

Nor "_two_different" or "_physical", although I suspect it's just a
terminology difference.  Both cores share the same die.

AMD dual cores processors are on a single die, with a single connection to
memory, and a single connection to hypertransport.  So the motherboard
sees no difference between a single core and a dual core.  There is
a small difference in the CPU initialization, so make sure your
motherboard/BIOS mentions support for dual core AMD chips.

Both cores are behind a system request interface (SRI), and a crossbar.

I wouldn't expect any problems from the newer RHEL or Fedora distributions.

Of course this only doubles performance (at the same clock speed) if your
code is relatively cache friendly, since your sharing a single memory
bus between 2 cores.  From my experience even memory intensive codes
often scale in the 1.5-1.8 times faster range.  Worst case, of course,
is 0% faster.  Actually the worst I've seen is around 5-10%, even on
completely memory bound code a dual core seems to manage to keep the
memory bus slightly busier.

-- 
Bill Broadley
Computational Science and Engineering
UC Davis