[Beowulf] Definition of HPC
prentice.bisbal at rutgers.edu
Tue Apr 16 13:11:47 PDT 2013
On 04/16/2013 03:50 PM, Max R. Dechantsreiter wrote:
> ...Perhaps HPC should be defined problem-down, rather than hardware-up.
> Regarding your "Issue #1:"
> There are some _nearly_ EP problems that require very minimal, but crucial,
> communication (perhaps at long intervals).
> As to your "Issue #2:"
> "Owned compute" has some advantages over "rented compute." In general, the
> control one has over one's owned resources enables applications to run with
> greater performance. Some optimizations just demand root access!
As someone who has been Scientific Computing/HPC System Admin, I can
tell you this is a complete myth. From my experience, the more access
users have to the hardware, the worse performance gets! In fact, in my
current job at a large university, I've already done my share of
cleaning up after people who had control over their owned resources.
And don't think this argument really applies to cloud computing, either.
It's my understanding that with AWS and other cloud services, you still
get root access to your virtualized guesses.
More information about the Beowulf