Artificial Intelligence Memory
At the 2019 Semicon Conference Applied Materials (AMAT) had a day-long seminar focused on technology, particularly memory, for artificial intelligence (AI) applications. In addition to talks by AI experts, the company also talked about their tools for manufacturing magnetic random access memory (MRAM) as well as resistive random access memory (RRAM) and Phase Change Memory (PCM). We will talk about a workshop at Stanford in August will explore emerging memories enabling artificial intelligence, especially for embedded products, such as IoT devices.
Gary Dickerson from Applied Materials gave a kick-off talk at the seminar. He talked about the growth of data and the importance of memory to support data centers as well as the edge. Increasing performance requirements will require improved throughput and latency for AI training in data centers (the Cloud) as well as inference engines in data centers and at the edge of networks.
MRAM, with its high performance and greater density potential than SRAM and other memory technologies, such as NOR flash, will be a key enabler of future edge inference technologies enabling sophisticated IoT devices and applications. RRAM and PCM will find their greatest applications in data centers. In particular, RRAM and PCM technology may be used in neuromorphic computing applications running sophisticated AI algorithms.
The image below shows Applied Material’s Endura Clover MRAM Physical Vapor Deposition (PVD) platform that can deposit the multi-level structures in STT-MRAM memory devices without going out of vacuum. The company also was showing their Endura Impulse PVD System for PCM and RRAM.
A one-day workshop at Stanford University on August 29, 2019, put on by the Stanford Center for Magnetic Nanotechnology and Coughlin Associates, features invited expert speakers to talk about various emerging non-volatile memories and how they will enable the next generation of artificial intelligence (AI) devices in the home, in the factory and in industry. The conference URL is: https://emai19.sites.stanford.edu.
Memories that will be discussed include Magnetic Random Access Memory (MRAM), Resistive RAM (RRAM), Ferroelectric RAM (FRAM) and Phase Change Memory (PCM). These memories are now available as discrete as well as embedded non-volatile memories and will be the basis of a new generation of devices, working in devices and on the network edge, as well as in data centers. These memories will enable AI inference engines that enable voice recognition, image recognition and system level optimization, all with minimal power utilization compared to conventional volatile memories.
The AMAT all day seminar at the 2019 Semicon showed the importance of emerging memories to enable AI applications in data centers and at the edge. A Stanford workshop in August will explore further the use of emerging memory in AI applications.