[THIN] Re: Compaq / HP Server testing results
- From: "Dez" <Dez_is@xxxxxxxxxxx>
- To: thin@xxxxxxxxxxxxx <thin@xxxxxxxxxxxxx>
- Date: Thu, 2 Jan 2003 17:44:31 +0000
Hi Adam, I read that stuff in the report on HyperThreading, havent really had
much of a chance to look into it yet to be honest will have a nosey around on
Intel's site tomorrow. The main thing that confused me was that they were only
claiming a 30 odd percent increase over a 380 with dual P3-1266 , and only 22%
over a dual P3-1400 DL380, which didnt seem a lot to me, as if they are all
comparable, it was basically saying a quad Xeon DL580 900Mhz was only 22% more
powerful than a dual P3-1400 DL380.
Chris
======== At 2003-01-02, 10:28:00 Adam.Baum@xxxxxxxxxxxxxx wrote: ========
I'm also looking at those reports. Part of the reason for the big jump is
that the DL380 G3 now uses Xeon processors that incorporate HyperThreading
HyperThreading sort of makes one CPU into 1.5 CPUs. I won't get into the
chip architecture, but if you are interested, Intel has quite a bit of info
on it
In comparing the quad 900mhz dl580 vs a dual 2.8Ghz Dl380, you get more
work done on the DL380. If you want to see how MS 32bit issues affect
things, check out the 1.6GHz DL580 report. It tops out at two processors,
not four due to PTE limitiations. If you want more, you'll need a 64bit
CPU/OS combination
adam
"Dez"
< Dez_is@xxxxxxx To: "thin@xxxxxxxxxxxxx" <
thin@xxxxxxxxxxxxx>
o.uk> cc:
Sent by: Subject: [THIN] Compaq / HP
Server testing results
thin-bounce@fre
elists.org
01/02/2003
09:30 AM
Please respond
to thin
Does anyone know how relevent the Compaq Thin Client server testing
documents are when comparing to the tests now done by HP? We currently have
a DL580 quad Xeon 700Mhz with 4Gb RAM, and certainly don't get near the
number of "heavy" users they claim to be able to sustain, but its the
comparison between one server and another that I'm more interested in,
rather than outright numbers, as I know our apps are heavy on the server
and don't expect to be able to replicate the ideal
Looking at the report on the DL580, it seems to suggest that a 900Mhz
P3-Xeon in QUAD CPU Configuration supports LESS users than a new DL380 G3
with Dual "Xeon" 2.8Ghz server in an otherwise similar configuration (120
"heavy" users compared to 122). The DL580 report was done by Compaq last
year, the new DL380 G3 report was done by HP, so I was wondering if the
testing was totally different and non-comparable? The reason I'm suspicious
is that the DL380 G3 report claims only a 34% performance hike over an old
DL380 1.266Ghz system, which doesnt sound like the kind of advantage a quad
DL580 should have over a dual P3-1.2Ghz based system
Chris
***********************************************
This Weeks Sponsor: 99point9.com
The 99Point9.com Online Tech Support
Helpdesk is the one-stop solution for all
your server-based computing needs
http://www.99point9.com
************************************************
For Archives, to Unsubscribe, Subscribe or
set Digest or Vacation mode use the below link
http://thethin.net/citrixlist.cfm
***********************************************
This Weeks Sponsor: 99point9.com
The 99Point9.com Online Tech Support
Helpdesk is the one-stop solution for all
your server-based computing needs.
http://www.99point9.com
************************************************
For Archives, to Unsubscribe, Subscribe or
set Digest or Vacation mode use the below link
http://thethin.net/citrixlist.cfm
= = = = = = = = = = = = = = = = = = = = = = = = = = = = =
***********************************************
This Weeks Sponsor: 99point9.com
The 99Point9.com Online Tech Support
Helpdesk is the one-stop solution for all
your server-based computing needs.
http://www.99point9.com
************************************************
For Archives, to Unsubscribe, Subscribe or
set Digest or Vacation mode use the below link.
http://thethin.net/citrixlist.cfm
Other related posts:
