3 ms·
Not enough that we’ve ever noticed it being at all significant in the profiling we do regularly. The system is an Edge ML processor running about 600 megapixel/
by tonyarkles 11mo ago
Not enough that we’ve ever noticed it being at all significant in the profiling we do regularly. The system is an Edge ML processor running about 600 megapixel/sec through a pipeline that does pre- and post-processing on multiple CPUs and does inference on the GPU using TensorRT. In general allocation isn’t anywhere near our bottleneck, it’s all of the image processing work that is.