[Beowulf] Cluster Diagram of 500 PC
john.hearns at streamline-computing.com
Fri Jul 13 01:58:53 PDT 2007
Mark Hahn wrote:
>> Anyway, hop latency in Ethernet is most of times just peanuts in terms of
>> latency compared to TCP/IP stack overhead...
> unfortunately - I'm still puzzled why we haven't seen any open,
> LAN-tuned non-TCP implementation that reduces the latency. it should be
> possible to do ~10 us vs ~40 for a typical MPI-over-Gb-TCP.
Well, the SCore impementation which we install on all our clusters does
In fact, we have one 500 machine cluster which (at the time of install)
ranked 167 in the Top 500 and achieved a very high efficiency.
All connected with gigabit ethernet only.
Senior HPC Engineer
The Innovation Centre, Warwick Technology Park,
Gallows Hill, Warwick CV34 6UW
Office: 01926 623130 Mobile: 07841 231235
More information about the Beowulf