This year’s AI Hardware Summit focuses on systems-level optimizations for specific AI workloads and results from novel AI acceleration technologies. We will be joining hundreds of other AI experts to discuss the future of AI. Find out more from the event agenda and tune in to our presentation.
- Co-designing systems that are fast, efficient, sustainable, and explainable.
- Optimizing systems for specific workloads like recommendation or NLP.
- Shortening time-to-value for ML models, from data through compute.
- Edge ML systems design and engineering.
- Novel ML techniques & architectures.
Presentation: Machines Like Me: See, Think, Do with Petascale at the Edge
Presenter: Andrew Grant, Senior Director, Artificial Intelligence at Imagination Technologies
Date/Time: Wednesday 15th September, 15:20 – 15:45 PDT
Abstract: Petascale AI at the Edge, as well as in the Datacenter presents challenges at the system level that can now be overcome. How custom build hardware systems can be flexibly configured from the key building blocks of
AI to achieve ultra-high performance via a heterogeneous systems approach. Software-driven, event-driven, data-driven but always ASIC hardware enabled. We then determine how this can be used flexibly within a number of
architectures to drive Robotaxis, as well as the industries of the future.