Skip To Main Content
Support Knowledge Base

What Is the Reason When Inferring FP32 Model Format Is Faster than FP16 Model Format on CPU?

Content Type: Compatibility   |   Article ID: 000093301   |   Last Reviewed: 11/20/2023
Description Resolution

Description

  1. Inferred FP32 model format and FP16 model format on CPU.
  2. Inference speed of FP32 model format was faster than FP16 model format.
  3. Unable to determine the reason when inferring FP32 model format is faster than FP16 model format on CPU.

Resolution

Inference result varies according to supported device plugin with supported model format.

Related Products

This article applies to 3 products.
Intel® Xeon Phi™ Processor Software OpenVINO™ toolkit Performance Libraries

Need more help?

Contact support
Contact support