Accurately estimating power for your vision SoC can make the difference between success and a multi-million dollar failure. Estimating power can be fairly straightforward for a RISC processor, but today’s vision SoC designs include neural networks with intense computation requirements making accurate power estimation much complicated. How can a designer have confidence in the power estimations that they rely on to deliver a product that meets the design requirements? Is the Quality of Results (QoR) representative for the end product power consumption? Applications that are used for power analysis along with different methods have an big impact on the quality of the results. Avoiding the pitfalls of inaccurate power estimation that are specific to vision SoCs can improve your SoC’s success on its way to market.

This presentation will cover:

  • The range of methodologies used for power estimation and verification
  • The advantages and disadvantages of each for AI-enabled vision SoCs
  • Typical ways some estimation tools falter when applied to vision neural networks
  • A proposed methodology to uncover the most accurate estimations from vendor selection through tapeout

Who should attend:

  • Design managers and system architects of vision-enabled SoC designs
  • Hardware design engineers

Derya Eker is an Engineering Manager at Synopsys with more than 20 years of semiconductor industry experience in digital design and verification of CPUs, IP and subsystems in various products and application domains spanning mobile communication, imaging, vision and neural networking. She has a relentless focus on quality, has been active in various global, multi-site technical forums to improve product quality and engineering best practices. At Synopsys, Derya leads a team that is responsible for defining and performing system-level verification across all ARC processor families. Prior to joining Synopsys, Derya held various management and engineering positions at Intel, Ericsson and NXP/Philips Semiconductors. Derya holds a bachelor’s degree in Electrical Engineering.