|
|
|
Figure 8
Benchmarking of TTCF computation using an arbitrary data set with array shape (300, 65536) and 10 q-bins, where 300 is the number of time points and 65536 = 512 × 128 is the reshaped dimension of a single AGIPD module. The `dense calc' refers to standard matrix multiplication using dense arrays. The `sparse calc' includes both the time required to convert the data to sparse format (`sparsification') and the subsequent multiplication of sparse matrices (`sparse math'). |

journal menu![[Figure 8]](zhu5014fig8.jpg)
access


