If you use a hub, the bandwidth is shared, meaning 10BT for two channels
(5BT for each channel). Each 10,000W (or 640K bits) array would require
0.064 x 2 + 0.000 which is still greater than 0.064 + 0.053. 0.053 is
obtained by assuming 1500-byte MTU and 0.001 sec latency (for each packet).
I also assumed that you have a no-latency net after using a hub.
yong
-----Original Message-----
From: R. M. Fesinmeyer [mailto:rmf_at_u.washington.edu]
Sent: Wednesday, October 03, 2001 11:37 PM
To: amber_at_heimdal.compchem.ucsf.edu
Subject: ?s: Xeon vs. Athlon, latency
Hello -- I have two questions I'm hoping can be answered:
1) Are any benchmarks numbers available for dual Athlon and dual (P4) Xeon
systems?
2) Network latency usually comes up whenever a conversation about PC-based
clusters start. In a previous conversation
(
http://amber.ch.ic.ac.uk/archive/) 100Mb/s
ethernet was frowned upon because of its high latency when using a switch.
My understanding is that standard hubs can have lower latency than the
average store-and-forward switch. For a 2-4 node (dual cpu) cluster, might
a hub provide better performance scaling?
With a very small cluster (2-4 nodes), it might even be better to do away
with the hub/switch entirely and connect each pair of systems with an
ethernet crossover cable (obviously this scales depending on your number of
PCI slots). Naturally this requires delving into some relatively-fancy
network set-up, but for the cost of 10Mb/s cards (I doubt amber could
saturate that connection between two nodes), it would seem like a very
inexpensive way of getting very-fast/low-latency interconnects. Is such a
possibility even worth considering?
Thanks for your input
Robert Fesinmeyer
Received on Thu Oct 04 2001 - 10:19:29 PDT
