Table C.1
Resource usage for various deconvolution runs.
| Name | l max | k max | T wall | T setup | T iter | N iter | Mem |
|
|
|||||||
| 30a | 400 | 6 | 10:07 | 1:13 | 3.9 | 140 | 6.6 |
| 30b | 500 | 6 | 25:49 | 1:20 | 5.5 | 270 | 8.1 |
| 30c | 600 | 4 | 1:24:57 | 1:07 | 5.1 | 995 | 6.4 |
| 30d | 600 | 6 | 1:08:53 | 1:35 | 6.9 | 591 | 9.6 |
| 30e | 600 | 8 | 1:40:40 | 2:24 | 10.3 | 573 | 11.9 |
| 30f | 700 | 6 | 3:41:13 | 1:48 | 8.5 | 1567 | 11.3 |
| 30g | 800 | 6 | 5:10:41 | 1:56 | 10.6 | 1759 | 12.8 |
| D | 500 | 30 | 10:31:38 | 8:34 | 20.3 | 1842 | 17.6 |
| 70a | 1700 | 6 | 63:51:56 | 15:27 | 78.3 | 2923 | 28.8 |
| 70x | 1700 | 6 | 9:43:05 | 5:58 | 11.8 | 2955 | – |
| 70y | 1700 | 6 | 5:25:32 | 7:20 | 6.6 | 2922 | – |
Notes. Twall denotes the total wall clock time of the run, Tsetup is the time required for the code startup (i.e. data input, computing the RHS, Wigner transform of the hit count and preconditioner), Titer gives the average wall clock time for a single CG step in seconds, Niter is the required number of iterations until convergence, and “Mem” is the total memory usage in GB. Please note the additional comments in the text.
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.