Single-binary interactive CLI harness to stress i.MX95 GPU/VPU (NPU later) and measure cross-block interference via DDR bandwidth
benchmark machine-learning arm cpp gpu v4l2 vpu opengl-es aarch64 soc neutron nxp embedded-linux memory-bandwidth tensorflow-lite npu edge-ai eiq imx95 mali-g310
-
Updated
Jun 13, 2026 - C++