Mixed Precision (Tensor) vs raw FP16 / raw FP32 Compute Metrics

Robotics & Edge Computing Jetson & Embedded Systems Jetson AGX Xavier
, , ,
5

Hi,

1.
No. It is low-level profiling, which is directly measured by half mode calculation.

2.
We don’t have an FP32 TFLOP score.
You can find detailed FP16 and INT8 compute data below:
https://developer.nvidia.com/blog/nvidia-jetson-agx-xavier-32-teraops-ai-robotics/?ncid=so-fac-mdjngxxrmllhml-69163

Thanks.