Implementation and Optimization of a CFD Solver Using Overlapped Meshes on Multiple MIC Coprocessors

Listing 2

: Memory arrangement on MIC.

(1)	!dec$ offload begin target(mic:n) in(mbc_n:length(len) alloc_if(.true) free_if(.false.)) …
(2)	!dec$ end offload
(3)	repeat
(4)	……
(5)	!dec$ offload target(mic:n) nocopy(mbc_n:length(len) alloc_if(.false.) free_if(.false.))
(6)	!$omp parallel do private (ik, i, k)
(7)	do ik = 1, ik_total
(8)
(9)
(10)	call Fluxj_mic(mbc_n, i, k, ……)
(11)	end do
(12)	!$omp end parallel do
(13)	……
(14)	until convergence
(15)	!dec$ offload begin target(mic:n) nocopy(mbc_n:length(len) alloc_if(.false.) free_if(.false.))
(16)	!dec$ end offload