feat: Intel GPU Max (Ponte Vecchio) OpenMP target offload support#1445

Draft

sbryngelson wants to merge 7 commits into

Commits on May 18, 2026

feat: Intel GPU Max (Ponte Vecchio) OpenMP target offload support
sbryngelson
committed
fix: replace integer kind literals with real literals in m_fftw.fpp
sbryngelson
committed
fix: extend VLA guards from USING_AMD to (USING_AMD or USING_INTEL) in m_qbmm and m_hyperelastic
sbryngelson
committed
fix: use shm:ofi + FI_PROVIDER=tcp for Intel MPI on crnch-gpu (tcp fabric removed in 2021.x)
sbryngelson
committed
docs: document Intel MPI multi-node SSH bootstrap workaround for missing bundled ssh
sbryngelson
committed
fix: add FI_PROVIDER_PATH to crnch-gpu modules; document SLURM GRES and OFI provider requirements
sbryngelson
committed
docs: document inter-node MPI fix (FI_TCP_IFACE) and dash3 renderD128 permission issue
sbryngelson
committed