SPRACZ2 August   2022 TDA4VM , TDA4VM-Q1

ADVANCE INFORMATION  

  1.   Abstract
  2. 1Introduction
    1. 1.1 Vision Analytics
    2. 1.2 End Equipments
    3. 1.3 Deep learning: State-of-the-art
  3. 2Embedded edge AI system: Design considerations
    1. 2.1 Processors for edge AI: Technology landscape
    2. 2.2 Edge AI with TI: Energy-efficient and Practical AI
      1. 2.2.1 TDA4VM processor architecture
        1. 2.2.1.1 Development platform
    3. 2.3 Software programming
  4. 3Industry standard performance and power benchmarking
    1. 3.1 MLPerf models
    2. 3.2 Performance and efficiency benchmarking
    3. 3.3 Comparison against other SoC Architectures
      1. 3.3.1 Benchmarking against GPU-based architectures
      2. 3.3.2 Benchmarking against FPGA based SoCs
      3. 3.3.3 Summary of competitive benchmarking
  5. 4Conclusion
  6.   Revision History
  7. 5References

3.3.3 Summary of competitive benchmarking

For deep learning performance and power benchmarking, it is essential to perform apples-to-apples comparison using industry standard machine learning benchmarks. MLPerf Inference [9] is an industry recognized standard for this purpose. Using these guidelines, we can illustrate the benefits of TDA4M edge AI processor resulting in greener and more efficient edge AI systems.

Different metrics used are:

  1. TOPS (Tera operations per second)
  2. FPS (Frames per second)
  3. FPS/TOPS: Frames per second normalized to 1 TOPS

TDA4VM platform’s advantage is summarized both from performance and power aspects as below:

  • GPU based architectures: TDA4VM offers up to 60% better performance efficiency measured as FPS/TOPS and 60% better power efficiency measured as FPS/Watts.
  • FPGA based architectures: TDA4VM offers more than six times performance boost in TOPS metric.