Open Access

Table B.1

Total PDR benchmark CPU time (in units of s) with different chemical solvers.

Model DGESV DGESVX DSGESV DLSODES
Fixed temperature models

F1 21 22 21 129
(n = 103, χ = 10) (24) (24) (23)
F2 34 36 33 131
(n = 103, χ = 105) (32) (35) (30)
F3 9 8 14 90
(n = 105.5, χ = 10) (10) (10) (10)
F4 10 9 9 210
(n = 1055, χ = 105) (11) (11) (11)
Variable temperature models

V1 17449 17090 ~ 3285
(n = 103, χ = 10) (466) (621) (492)
V2 1897 4816 3371 3743
(n = 103, χ = 105) (514) (645) (599)
V3 1524 2488 4414 9803
(n = 1055, χ = 10) (1686) (3026) (1411)
V4 1664 1912 1812 23592
(n = 1055, χ = 105) (179) (1303) (1489)

Notes. The numbers in parenthesis give the results for runs where a time-dependent solver was used as backup approach after the stationary solver failed. The last column corresponds to runs where DLSODES was immediately used instead of a steady-state solver. (~) indicates jobs stopped after 60 iterations. All computations were done on a 8-core Intel(R) Xeon(R) CPU E5-2630 v3 2.40GHz with 64 GB memory. No parallelization was applied. Densities n are given in units of cm−3.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.