[Beowulf] Acceptable rad limits for cluster rooms?

[Greg Lindahl]

On Fri, Jun 16, 2006 at 03:21:56PM -0600, Brian Oborn wrote:

> The cluster for our Physics department is next to a room that, at the 
> time of installation, was an empty accelerator hall.

I'd test it. If you can find the count of single-bit upsets, even one
machine for a week or two will give you plenty of confidence in
the number of 2-bit errors for a cluster.

-- greg