Στις Παρ, 10 Ιαν 2020 στις 12:41 μ.μ., ο/η Ioannis Venetis <ioannisvenetis@gmail.com> έγραψε:
BTW just three days ago eigensym was switched to a different Lapack
background function I think, so maybe we should try a brand new snapshot
for this.


It could always be that openblas has different optimizations for
different Lapack functions for your newer CPU.

cheers
sven


Using 2020a-git (2020-01-08) returns the same times. Also I run the script from either the desktop location or from the hard disk. No other program is running etc.

I attach my $sysinfo

? eval $sysinfo
bundle anonymous:
  nproc = 12
  blascore = "Haswell"
  hostname = "DESKTOP-OFGVHSS"
  os = "windows"
  mpi = 0
  blas = "openblas"
  omp_num_threads = 12
  ncores = 6
  omp = 1
  blas_parallel = "OpenMP"
  mpimax = 12
  wordlen = 64

if it could be of any value

Yiannis

OK I went back to my old PC (I have 2 I admit!!) and I use 2019d in both cases.
I get a ratio of 10 (less than half). But 'eigensym' does almost the same time while eiggen2 is much slower.

So in my new (and I guess better PC) with eigensym I get the same timeA but a much faster timeB from 'eiggen2'.

Riddle to be solved …  :).

Same thing happens with 2020a-git in both machines

Thanks
Yiannis