Latency Performance Comparisons - 2023.2 English

Vitis Tutorials: AI Engine (XD100)

Document ID
Release Date
2023.2 English

Following is a table comparing the executions times to simulate 12,800 particles for one timestep on the different N-Body Simulators explored in this tutorial.

Name Hardware Algorithm Average Execution Time for 1 Timestep (seconds)
Python NBody Simulator x86 Linux Machine O(N) 14.96
C++ NBody Simulator A72 Embedded Arm Processor O(N2) 123.299
AI Engine NBody Simulator Versal AI Engine IP O(N) 0.007

As you can see, the N-Body Simulator implemented on the AI Engine offers a x2,800 improvement over the Python O(N) implementation and a x24,800 improvement over the C++ O(N2) implementation. A vectorized C++ NBody Simulator O(N) implementation can be created with pthreads, but is left as an exercise for the user.