search

ValeevGroup / tiledarray

A massively-parallel, block-sparse tensor framework written in C++
GNU General Public License v3.0
247 stars 51 forks source link

cmake 3.28+: CUDA support broken #438

Closed evaleev closed 7 months ago

evaleev commented 7 months ago

in cmake 3.28 FindCUDA (used by BLT via Umpire) requires policy 146 set to OLD

image

revamp of CUDA support in BLT is not yet ready: https://github.com/LLNL/blt/pull/585