David Williams-Young
David Williams-Young
> So A.clone() would create a deep copy of A. Yes, I suppose the following syntactic sugar would be ``` Matrix clone() { auto cpy = this->emptyLike(); cpy.insertLocalTiles(); copy( *this,...
> Thinking more on this This is essentially what TiledArray does (by introducing the concept of `TiledRange`), however, the individual tiles also carry information pertaining to their global location (i.e....
This is partially addressed in #111 - GPU code directly evaluates the Laplacian rather than the Hessian
@susilehtola In prep for discussion with CRYSTAL folks, can you drop some references here?
Hi @NVIDIA-JerryChen The notion of separating out the spatial Rho construction is something I've considered in other contexts (PGAS, incremental XC, etc), but I think it would be useful here...
@evaleev I don't doubt we're going to have leakage, and I'm totally fine with that. We need this to be as close to hardware as possible. If someone can get...
@loriab I can see a few ways we could fix this: 1. Add a shell ordering to `Shell` - This would allow for the most flexibility (mixing ordering if one...
@ryanmrichard These are all good points. A few clarifications: 1. I agree that downstream users should pin GauXC revisions. However, there are a number of projects that we integrate with...
@susilehtola would an API to return a _**specific**_ atomic overlap (as opposed to all or a batch of them) suit the needs here (at least to start)? Would it also...
Also, do you have a reference code for this? Code is written, but need something to verify against