[Swift-devel] VDS1 transfer executable

[Ben Clifford]

On Tue, 24 Jul 2007, Ian Foster wrote:

> Mike mentioned that he thought Nika's MolDyn code had some workaround in 
> it to reduce data movement, introduced because of a lack of caching 
> support. Does anyone know about that?

that is bug 76: http://bugzilla.mcs.anl.gov/swift/show_bug.cgi?id=76

Bug 78 http://bugzilla.mcs.anl.gov/swift/show_bug.cgi?id=78 (or rather, 
the privately discussed rather different root cause of bug 78, which is to 
access dcache data) is higher priority.

A basic approach is to have the submit side access dcache, as I've 
discussed elsewhere; there's no direct evidence that that approach will be 
unsuitable (though thoughts that it might be are what motivated this 
thread).

We can look at doing that next.

--