Hans Pabst

Pretty naive, but #4504 wants to add the total time to the title (not changing the table or graph).

I will go over all of the above regressions/claims. Regarding CMake-built CP2K, I verified the Spack builds are capable of matching performance levels of legacy builds with most deps linked...

Further, we can merge #4509. I am conducting some final tests to reproduce the original motivation (less memory consumption of page-locked memory). For the sake of getting back performance for...

Regarding multiple threads, as a fair measure - shall we use `OMP_PROC_BIND=true`? It does not make too many assumptions like too sophisticated.

Let's close the issue as completed if there are no more pressing issues.

For the record, this was your call when the Regtest were sorted out. Now you're doing it yourself; much appreciated!

> It doesn't look too good: While the ifort [gets stuck](https://storage.googleapis.com/cp2k-ci/run-cp2k-intel-ifort-psmp-8d7e2530_report.txt) on [grid_dgemm_prepare_pab.c](https://github.com/cp2k/cp2k/blob/master/src/grid/dgemm/grid_dgemm_prepare_pab.c) the ifx [struggles to parse DBCSR](https://storage.googleapis.com/cp2k-ci/run-cp2k-intel-ifx-psmp-7261e4db_report.txt). I think, I prepared DBCSR for Intel Compiler and had various...

Updating from oneapi-hpckit:2025.2.2-0 to oneapi-hpckit:2025.3.0-0 is not a good idea since IFX 2025.3 regressed (I am on it). See also https://github.com/cp2k/cp2k/pull/4553#issuecomment-3574219832.

> I rebased onto DBCSR 2.9.0 and went back to `oneapi-hpckit:2025.2.2-0`. Unfortuantely, not much has changed: > > * ifort still [crashes](https://storage.googleapis.com/cp2k-ci/run-cp2k-intel-ifort-ssmp-fea5b11a_report.txt) > > * ifx fails with a [linker...

That's fine too. Once we have migrated, I can also enable some currently unsupported combinations. For instance, GNU Fortran and Intel MPI/MKL are a good "team". It's my own preferred...