|
|
## Changelog for upcoming release ##
|
|
|
## Changelog for next release ##
|
|
|
|
|
|
- not yet decided
|
|
|
|
|
|
Changelog for upcoming ELPA 2021.11.001.rc1
|
|
|
Changelog for ELPA 2021.11.001
|
|
|
|
|
|
- support of Nvidia cusolver library to accelerate some routines (needs CUDA >= 11.4)
|
|
|
- experimental Nvidia GPU versions for "elpa_invert_trm" and "elpa_cholesky"
|
... | ... | @@ -12,6 +12,9 @@ Changelog for upcoming ELPA 2021.11.001.rc1 |
|
|
- allow to call ELPA eigenvectors and eigenvalues also with GPU device
|
|
|
pointers for the input matrix, the vectors of eigenvalues and the output
|
|
|
matrix for the eigenvectors
|
|
|
- EXPERIMENTAL feature:g new real GPU kernel for Nvidia A100 (provided by Nvidia): can show a
|
|
|
performance boost if number of vectors per MPI task is > 20000. Most likely
|
|
|
most benifit in non-MPI version
|
|
|
- as anounced, droping the legacy interface
|
|
|
- more autotuning features, for example using non blocking MPI collectives
|
|
|
- new version of autotunig avoiding a combinatorial grow of possibilities
|
... | ... | @@ -20,7 +23,7 @@ Changelog for upcoming ELPA 2021.11.001.rc1 |
|
|
20211125)
|
|
|
|
|
|
Changelog for ELPA 2021.05.002
|
|
|
- no feature changes for users
|
|
|
- no feature changes
|
|
|
- correct the SO version which was wrong in ELPA 2021.05.001
|
|
|
|
|
|
Changelog for ELPA 2021.05.001
|
... | ... | |