- 14 Oct, 2015 2 commits
-
-
Edgar Solomonik authored
Turned off lapack routines when not building with -DTUNE
-
Edgar Solomonik authored
Setup configure file to make the config.mk file better documented and more easily modifiable to enable GPU support
-
- 12 Oct, 2015 2 commits
-
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
- 09 Oct, 2015 3 commits
-
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
got rid of printfs and gave up on trying to find actual memory used, instead just keeping track of tensor data stored and setting memcap to half the memory
-
- 05 Oct, 2015 1 commit
-
-
Edgar Solomonik authored
Implemented available memory check via malloc_info instead of mallinfo, which is not able to handle 64-bit
-
- 02 Oct, 2015 12 commits
-
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
Also replaced two unrelated stack with heap allocs.
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
- 01 Oct, 2015 1 commit
-
-
Edgar Solomonik authored
-
- 22 Sep, 2015 3 commits
-
-
Edgar Solomonik authored
-
Edgar Solomonik authored
Combined the old and new contraction mapping mechanisms and select mapping with lowest time among the two sets.
-
Edgar Solomonik authored
made other corrections to get new mapping scheme working properly
-
- 21 Sep, 2015 5 commits
-
-
Edgar Solomonik authored
Finished a seemingly working implementation of a more exhaustive mapping scheme, but performance seems like there are still some issues.
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
- 18 Sep, 2015 3 commits
-
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
Edgar Solomonik authored
Added model output/input functionality and setup initial models based on running CCSD on a 12-core linux machine
-
- 17 Sep, 2015 3 commits
-
-
Edgar Solomonik authored
Could use a bit more correctness verification.
-
Edgar Solomonik authored
Fixed correctness bug involving different processors folding differently due to differing performance models. Fix is temporary, simply broadcasts processor 0s solution. Planned solution is to do flat-tree TSQR.
-
Edgar Solomonik authored
-
- 16 Sep, 2015 3 commits
-
-
Edgar Solomonik authored
Still seeing a bug for sparse_mp3 on 24 processes.
-
Edgar Solomonik authored
-
Edgar Solomonik authored
-
- 15 Sep, 2015 2 commits
-
-
Edgar Solomonik authored
Made all bcast and allreduce calls of nontrivial size use the wrapped MPI routine through CommData, where times get measured and contribute to time estimation.
-
Edgar Solomonik authored
-