The AI Hardware Summit wrapped up on 15 September, 2021 at the Computer History Museum in Mountain View, CA. The summit is THE place for those accelerating AI workloads in at the cloud and at the edge, at both a hardware and systems level.
Before the event, Siemens hosted a live webinar, “Building AI Chips that Deliver” with an expert panel in discussion of the next challenges that lie ahead in the deployment of AI and ML SoCs. That webinar is available to watch on demand.
Register and watch the webinar “Building AI Chips that Deliver”.
Gajinder Panesar, Siemens Digital Industries Software Fellow, presented the case for on-chip analytics for AI hardware both in design and in production to solve tricky development and validation problems. With direct examples drawn from customer experience, he examined how AI chip designs can be more quickly and efficiently validated, debugged, deployed and optimized, both before market launch and during the full lifetime of the device. The recording of his presentation, “Functional analytics: the need for system-level visibility,” is available now for anyone to watch.
Watch the video presentation of “Functional analytics: the need for system-level visibility.”
Embedded Analytics Demonstration:
Meanwhile, in the Siemens exhibitor booth, we demonstrated how the Tessent Embedded Analytics silicon IP and software tools give unparalleled visibility and insights into complex functional behavior of AI SoCs. We showed how to collect relevant functional data from the chip using Embedded Analytics on-chip monitors and how to explore and analyze the data to glean unique operational insights that would otherwise be out of reach. The demonstration was recorded and is now available for anyone to watch.
Watch the technical Tessent Embedded Analytics demo.
Embedded Analytics for AI technical paper:
For more information on optimizing AI systems with Embedded Analytics, download the technical paper:
Harness system-level data to optimize manycore AI and ML chips | Siemens Digital Industries Software