How AI is Redefining Semiconductor Design for the Edge
Enabling Intelligence Through Sustainable Computing
The exponential growth in compute demands of AI, and the move to Software Defined Products, means that more than ever workloads are defining the semiconductor requirements. The need to hit the restrictive power, performance, area and cost constraints of edge designs, mean that every element of the design needs to be optimized and co-designed with the workloads in mind. Additionally, the design needs to evolve even after semiconductor design is complete as it adapts to new demands.
In this presentation, we will look at how semiconductor design is changing to enable rapid development and deployment of custom, application-optimized, system-on-chip designs – from concept through to in-life operation, as we chart the path to a sustainable compute future.
(AI HW Summit Keynote Delivered September 10th 2024, San Jose, CA)
Related resources
Catapult AI NN – From AI/ML Framework to Optimized RTL
Video describing how Catapult AI NN now delivers a methodology and flow from “AI/ML Framework to RTL” enabling rapid exploration of network, quantization, design reuse factors, and more!
LG Electronics: Video Encoder IP Design Optimization and Verification Using the Catapult Platform
Through a new design approach, LGE's SoC Center concluded that adopting Catapult in IP development significantly increases efficiency in terms of time and cost, and plan to increase Catapult use in future IP development projects.
NVIDIA: High-Level Synthesis in Agile System-on-Chip Flows: Overview and Techniques
This talk provides a brief overview of NVIDIA Research’s use of Catapult HLS and highlights some useful features and flows of the Connections library, such as the ability to back-annotate SystemC simulations.