Implementation of an Agent-Based Parallel Tissue Modelling Framework for the Intel MIC Architecture

Table 1

OpenMP scalability of two stages of simulations in Timothy.

Threads per node	Computing potential seconds (speedup)	Computing gradient seconds (speedup)
1	57.75	64.02
2	29.91 (1.93x)	37.73 (1.70x)
4	16.27 (3.55x)	22.55 (2.84x)
8	8.64 (6.68x)	12.11 (5.29x)
16	4.85 (11.91x)	6.59 (9.71x)
32 (SMT2)	2.90 (19.91x)	3.88 (16.5x)
64 (SMT4)	1.99 (29.02x)	2.78 (23.03x)