D. A. Clarke
D. A. Clarke
in gaugeActionDeriv.h: there is a claim symanzikGaugeActionDeriv is slower than gauge_force?. it seems like the only difference is one uses gsiteMu while the other uses both gsite and mu... dennis...
right now it initializes a `stored_checksum` to zero. its `read_header` method then never touches it. then in `readconf_milc_host`, it does a checksum comparison... well of course that will always fail,...
It is not written in a way that is general, readable, well-organized, or in some cases correct. For example: - There are parts which rather pertain to LIME packing generally...
this section of the code doesn't really use anything in our paradigm. the most egregious problem is that, in order to write out, it allocates memory for a buffer on...
It seems that the half-precision stuff doesn't compile for GTX Titan X cards, architecture=52. In particular it doesn't seem to like the half precision stuff. One option could be to...
Since testing.h a header file, I guess it would make a little more sense to me if it were in `base` or something. Furthermore, it would be nice if the...
- [x] Formatting makes hisq smearing and force modules and associated files difficult to read. Should be cleaned up. - [ ] DSlash has `HisqMdaggMFunctor` which is not used anywhere?...
there seems to be something wrong with milc to nersc conversion on multiple GPUs
Would be nice if this class - [ ] Had a way to get back correlators indexed by displacement vectors instead of the magnitude - [ ] Had way to...
1. It would be useful to add a short test to the CommunicationBase setup. This way with each run we can be sure e.g. cuda-aware mpi and cuda p2p are...