Exascalers, Hyperscalers, Computing at the Edge: AI Semiconductors Pull from Different Compute Modes and From Different Process Nodes
Can a machine learn to write for The New Yorker? According to John Seabrook, in the October 14, 2019 issue of the magazine, “Extraordinary advances in machine learning in recent years have resulted in A.I.s that can write for you.”
Can an edge-machine always be listening to you, always sensing you, always responding to you, on a long-term, low-power-draw basis? A company called Aspinity thinks so; its “Ultra-low-power analog signal processor, the RAMP chip, identifies voice at the earliest point in the audio signal chain, directly from the raw analog microphone data. Just as our brain selects only the important sounds to send deeper into the brain for more processing, the RAMP processor keeps the ADC, DSP, and other heavy power-consumers further down the audio chain either off or in a low-power mode — unless voice has actually been detected.”
Can we split the difference between the semiconductor process needs of a digital, trillion-transistor, wafer-scale AI chip, the Wafer-Scale-Engine from Cerebras Systems, and the corner of the semiconductor world where Aspinity (and its Reconfigurable Analog Modular Processor, RAMP, a “small-footprint ultra-low power analog neuromorphic processor”), or Syntiant (and its Neural Decision Processors, with their “highly accurate wakeword detection in a tiny package with near-zero power consumption”) operate?
Can we productively have a foot in both camps, the semiconductor fabrication industry’s right foot serving the needs of Hyperscaler customers (IBM, Cerebras, Tesla, and the FAANGS of the world) with their Exascaler compute (computing systems capable of at least one exaFLOPS, or a billion billion calculations per second) aspirations, and simultaneously have our left foot manufacturing smart, compact, analog-based sensing circuits in die volumes orders of magnitude larger than the flagship trillion-transistor-race products run?
Does “Intel’s billion-dollar transformation in the age of AI” augur a sea change in the semiconductor manufacturing world?
Please join us for the ASMC 2020 Panel Discussion, in Saratoga Springs, New York, the afternoon of Wednesday 06 May 2020, where distinguished panelists drawn from the AI, Exascaler, Hyperscaler, and Edge Computing communities will offer us their learned comments in a discussion with the audience on just what it is these companies need and expect from the advanced semiconductor manufacturing world, and what that means to the SEMI ASMC 2020 attendees as we deliver silicon for the AI-powered future of the 2020s, and beyond.
Moderator: Paul Werbaneth, Guest Editor, 3D InCites
- Buvna Ayyagari-Sangamalli, Head of Design Technology, New Markets and Alliances Group, Applied Materials
- Bahgat Sammakia, PhD, VP of Research, Director, Center for Energy-Smart Electronic Systems, Binghamton University
- Xavier Lafosse, PhD, Commercial Technology Director, Advanced Optics, Corning
- Anna Topol, PhD, Chief Technology Officer, IBM Global Think Labs
- Carl P. Evans III, Founder, Tercero Technologies