General

Profile

Peter Kasson

Issues

Projects

Activity

07/20/2017

07:05 PM GROMACS Revision 54821ced: Rework and rename gmx_select_gpu_ids
This change makes clear that this function now produces a validated
mapping of intra-node PP ranks (ie with GPU tasks...

03/09/2017

07:30 PM GROMACS Task #2045: API design and language bindings
Design work is ongoing. Eventually a draft API spec and before that a simple usage diagram will come here.
We're st...

03/08/2017

03:10 AM GROMACS Task #1781: re-design benchmarking functionality
I'll generally stay out of the...discussion about what should be done in this particular case, but I think that as a ...

02/20/2017

07:29 PM GROMACS Feature #1137: Proposal for integrator framework (do_md) in future GROMACS
Maybe we want to either break the StateManager discussion into another thread or reserve detailed consideration for l...

10/29/2016

05:21 PM GROMACS Feature #2068: Access to low level classes
Would it be helpful to differentiate between modifying settings "in flight" i.e. during a simulation and modifying se...

08/31/2016

05:47 PM GROMACS Task #2045 (New): API design and language bindings
We're gearing up for external API design. We ultimately want language support in Python and a C/C++ language. One q...

07/03/2016

01:31 PM GROMACS Feature #1972: external potential modules for refinement against experimental data
At the meeting we had, we put a starting point spec on the board that everyone there (Erik, David, Berk, Mark, Christ...

07/02/2016

10:01 PM GROMACS Feature #1972: external potential modules for refinement against experimental data
As we discussed in the meeting, I *strongly* believe we need the API specified (i.e. the external interface) and kept...

05/26/2016

12:37 AM GROMACS Task #1971: Removing buggy features vs. keeping workflows
I would argue that there should be another category. If saying that a feature is important requires a commitment by ...

05/25/2016

02:11 PM GROMACS Bug #942: -deffnm with enforced rotation code writes files that break checkpoint resume
I would *strongly* oppose removing -deffnm. It's an option we use all the time. For our usage I don't care if rotat...

Also available in: Atom