Hello Bob,

Re: TCP Checksum performance

Thanks for this feed back on the 937SX and 917LX.  I am a HP guy
and should know this, but my true love is MPE networking (don't
tell my wife) and with the huge number of CPU, DISK, PRINTERS,
etc. that are found on the 3000 the numbers after a while they
all just blur together, or it could just be my short term memory
lost (me not the 3000).  Actually it must be the serious lack of
memory in my 917LX that makes it so slow, it has always run like
a dog for me... which is not a bad thing since I do a bunch of
resource testing between fast and slow machines and when I need
a slow destination machine I know where to look.

I found the URL:

  http://raven.utc.edu/cgi-bin/WA.EXE?A2=ind9502B&L=hp3000-l&P=R7835

           Perf.
CPU        Rating
917LX       10.0
...
937         10.0
...
967         20.0
...

I did not find a ranking for my 959/200.

Thanks for the info, and as I warned YMWV - Your Mileage Will Vary.

Regards,

James Hofmeister
Hewlett Packard
Worldwide Technology Network Expert Center
P.S. My Ideals are my own, not necessarily my employers.


-----Original Message-----
From: Bob J. [mailto:[log in to unmask]]
Sent: Tuesday, August 15, 2000 15:09
To: HOFMEISTER,JAMES (HP-USA,ex1)
Subject: Re: [HP3000-L] TCP Checksum performance


James,
  The measurements are interesting but FYI a 937SX is not a faster
machine than a 917LX. The only differences between the 917, 927,
937, 947, LX, RX, and SX are licensing and numbers of I/O slots and
internal peripheral space with the required power supply(ies) to
support the load. They all run at exactly the same processor speed. If
you
want verification check a service manual or ask a CE.
  I'm going to turn my checksumming ON. From the info you provided it
sure seems worth the tiny performance degradation.

                     Thanx for the info.

                     Bob J. -- Ideal Computer Services



">
> The finding is on my low end 917LX is that data-transfer performance
> with TCP checksum enabled was significantly slower in the range of 7-8
> percent slower, but the CPU impact was only in the range of 2-3 percent
> higher overhead.
>
> My finding on my next fastest machine a 937SX is