Processor contention(?) and network bandwidth on AMD

[Joshua Baker-LePain]

On Mon, 29 Apr 2002 at 5:07pm, Robert G. Brown wrote

> Try this:
> 
>   Do your netpipe transmitter on an unloaded host, a host at load 1 and
> at load 2.
>   Do your netpipe receiver on an unloaded host, a host at load 1 and one
> at load 2.
> 
> Fill in the matrix -- load 0 to load 0, load 0 to load 1, etc.
> 
> I found (in similar tests done years ago) that a TRANSMITTER could be
> loaded to 2 (per cpu) with only a small degradation of throughput, but
> loading a RECEIVER would drop throughput dramatically, by as much as
> 50%.

Done -- the results are at <http://www.duke.edu/~jlb17/bwtest.pdf>.  Keep 
in mind these were pretty quick and dirty tests using the systems I have 
on hand.

Between Athlon systems, it seems the transmitter vs. receiver loading 
doesn't make much of a difference.  A newer Intel based system (dual PIIIs 
on a Serverworks HE-SL chipset) shows the same bandwidth hit with overload 
as the Athlon systems.  But older systems (well, a couple of years 
anyway) don't show this hit, which is what set me off on all this.  Maybe 
it's an issue of chipset support?

Anyways, thanks for listening to me babble on about all this.

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University