[Beowulf] Servers Too Hot? Intel Recommends a Luxurious Oil Bath

Jeff Johnson jeff.johnson at aeoncomputing.com
Wed Sep 5 08:46:27 PDT 2012

On 9/5/12 7:23 AM, Joe Landman wrote:
> In HFT latency matters.  Really ... seriously ... matters.  36
> milliseconds to get to the DC?  Won't fly.  The folks we work with are
> looking to shave microseconds off everything.  Anything and everything
> is on the table for this.

This is why it will never be anything but Secaucus and Carteret no 
matter what the cost for power and cooling. The HFT guys make 10X+, in 
seconds, the cost of a high PUE by having the lowest latency to the 
exchanges. When they are paying NRE to have FPGAs with trading rules 
built into 10GbE switches sitting at the network demarcation boundary 
things like cheap power and cooling mean little.

Jeff Johnson
Aeon Computing

jeff.johnson at aeoncomputing.com
t: 858-412-3810 x101   f: 858-412-3845
m: 619-204-9061

/* New Address */
4170 Morena Boulevard, Suite D - San Diego, CA 92117

More information about the Beowulf mailing list