Working to the deadline of Workunits

eOn code for long time scale dynamics

Moderator: moderators

Post Reply
ScouseSandwich
Posts: 5
Joined: Tue Nov 16, 2010 8:53 pm

Working to the deadline of Workunits

Post by ScouseSandwich »

Hello,

I have read about these forums that (currently) work units quickly become irrelevant. This is because saddle point calculations (currently) are very short and to get enough to satisfy your requirements does not take long.

So this brings me to ask, if I complete work units shortly before their deadline, are they likely to be irrelevant? this tend to be the case for me.

EDITED a couple of time to get rid of mistakes, shouldn't have rushed it!
chill
Posts: 96
Joined: Tue Jul 28, 2009 9:04 pm

Re: Working to the deadline of Workunits

Post by chill »

ScouseSandwich wrote:Hello,

I have read about these forums that (currently) work units quickly become irrelevant. This is because saddle point calculations (currently) are very short and to get enough to satisfy your requirements does not take long.

So this brings me to ask, if I complete work units shortly before their deadline, are they likely to be irrelevant? this tend to be the case for me.

EDITED a couple of time to get rid of mistakes, shouldn't have rushed it!
We have a new type of job we are running (parallel replica dynamics) and for these jobs the workunits are useful for longer periods of time. So if you complete your work unit closer to the deadline it is still useful. That isn't to say that we are no longer running saddle searches. We just have an additional type that we are running now.
Post Reply