Skip to Content.
Sympa Menu

forum - Re: Re: [abinit-forum] Using remote computers

forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: Re: [abinit-forum] Using remote computers


Chronological Thread 
  • From: <jimprtlnd01@gmail.com>
  • To: anglade@gmail.com, forum@abinit.org
  • Subject: Re: Re: [abinit-forum] Using remote computers
  • Date: Tue, 21 Oct 2008 21:10:18 +0200 (CEST)

I have access to some computers that are not local, but often have idle
processors. Additionally, on request I can get these computers dedicated to a
run.

I am running on 12 local processors. I would be able to add 4, or maybe more,
processors to a run. I am doing large molecules, with high complexity, so I
was assuming that the message/data passing is pretty insignificant compared to
the message passing. For example, crystals with 3-5th row elements, such as
copper, iron, calcium, barium, etc.

However, I am unable to benchmark or even try this because it just hangs upon
execution. And like I said previously, it works correctly, the same exact
way,
if the computers are all in-house.

If execution over the internet becomes impractical, could you please explain
why a little more? Is the message passing and memory allocation and all that
really such a problem? All computers are connected to broadband internet
connectiions, 1.5M or higher.

Thanks for helping me to understand.

Jim



Archive powered by MHonArc 2.6.15.

Top of Page