lkml.org 
[lkml]   [1998]   [Jun]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: quad ppro Compaq proliant 5000 problems with 2.1.106-ac4 was: Re: Any SMP people out there with SCSI CD ROMs?

On Thu, 25 Jun 1998, Joel Jaeggli wrote:

> On Thu, 25 Jun 1998, Chris Pirih wrote:
>
> > >> > Robert HYATT wrote:
> > >> > > I ran the matrix multiply benchmark... with 1, 2, 3 and 4
> > >> > > processors.
> > >> > > The numbers I got were 52 seconds, 59 seconds, 66 seconds and 73 seconds
>
> At least you know matrix multiply is totaly blocked on main memory,
> otherwise you might see 1/4 the performance with 4 processes.
>
> > Look everybody, it's very simple. 1 copy of the program takes 52
> > seconds. 2 copies of the same program running at the same time take
> > 59 seconds each. 3 copies take 66 seconds, 4 take 73. Each process
> > runs on its own CPU, but when several of them are running each takes
> > a little longer because they contend for the slow main memory, even
> > though each process has a CPU all to itself.
>
> This is one of the drawbacks with ppro's is that their chipsets,the 450gx
> and 440fx have about half the memory bandwidth of the p2's lx chipset with
> the bx chipset providing slightly higher throughput than that. Of course
> 450gx is the only way to go for 4 processors on a pc (for the next couple
> of days anyway) until the 450nx and the slot2 p2 make their entrance. see
> the discussion about this in the beowulf list about 6 months ago.


This Orion chipset isn't half bad, using 4-way interleaving to keep up
with the 4 processors. It scales pretty well when compared to some dual
PII machines I have tested that *really* bog down when you run two of
these things. I was surprised as this is a 66mhz bus also.


>
> http://cesdis1.gsfc.nasa.gov/hypermail/beowulf/beowulf.9712/subject.html#start
>
> or more specifically:
>
> http://cesdis1.gsfc.nasa.gov/hypermail/beowulf/beowulf.9712/0067.html
>
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.rutgers.edu
> >
>
> --------------------------------------------------------------------------
> Joel Jaeggli joelja@darkwing.uoregon.edu
> Academic User Services consult@gladstone.uoregon.edu
> PGP Key Fingerprint: 1DE9 8FCA 51FB 4195 B42A 9C32 A30D 121E
> --------------------------------------------------------------------------
> It is clear that the arm of criticism cannot replace the criticism of
> arms. Karl Marx -- Introduction to the critique of Hegel's Philosophy of
> the right, 1843.
>
>


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:43    [from the cache]
©2003-2011 Jasper Spaans