
The Massachusetts Institute of Technology (MIT) has launched an Artificial Intelligence (AI) programme to bring together its research on hardware with five industrial partners.
MIT has established the programme to coordinate various research groups at MIT working on AI. A key aim is to apply hardware technologies to improve AI engines in performance and energy consumption.
The five inaugural members of the programme are Amazon, Analog Devices (ADI), European chip making equipment developer ASML, US-based NTT Research and TSMC, the world’s largest chip making foundry,
The programme is looking at a wide range of AI hardware, from analog neural networks, in-memory computing and new roadmap CMOS designs to quantum AI and software-hardware co-design. It will also include AI edge security, wireless technologies, hybrid-cloud computing and high-performance computing (HPC).
NTT’s Physics and Informatics (PHI) Lab in Sunnyvale, California, on existing research as well as CMOS roadmaps, monolithic 3D systems, analog non-volatile memory devices and new concepts.
It will also include areas of particular interest, such as edge AI applications and integrated intelligent sensors, with a focus on scalability and energy consumption.
Related articles
The computational requirements for the machine learning subset of AI, for instance, are accelerating rapidly, especially where neural networks are concerned. Such a surge in computational requirements also has an impact on environmental sustainability, as the energy consumption grows rapidly to support such computational power. To address these issues, promising solutions will require attending to hardware, and perhaps revolutionizing components and architectures.
“A sharp focus on AI hardware manufacturing, research, and design is critical to meet the demands of the world’s evolving devices, architectures, and systems,” said Anantha Chandrakasan, dean of the MIT School of Engineering and Vannevar Bush Professor of Electrical Engineering and Computer Science. “Knowledge-sharing between industry and academia is imperative to the future of high-performance computing.”
“Expanding our relationship with MIT through this new program makes good sense, as it builds upon our existing and productive engagements, which have generated two dozen papers and several patents in less than three years, and makes available additional access to MIT’s extensive research activities in AI, providing further opportunities for collaboration,” said Kazuhiro Gomi, President & CEO of NTT Research, launched in July 2019 as a subsidiary of Japanese telecoms giant NTT.
NTT Research has entered a similar industrial partnership with the Simons Institute for the Theory of Computing in California and the PHI lab works with nine other universities, including the California Institute of Technology (CalTech), Cornell, Harvard University, Notre Dame University, Stanford University, Swinburne University of Technology, the Tokyo Institute of Technology, the University of Michigan, and the University of Tokyo. The NASA Ames Research Centre in Silicon Valley and 1QBit, a private quantum computing software company, have also entered joint research agreements with the PHI Lab.
The new program is co-led by Jesús del Alamo and Aude Oliva, and Anantha Chandrakasan serves as chair and NTT Research has joined for a three-year term that began in January 2022.
www.mit.edu; www.ntt-research.com
Related articles
- NTT buys Fujitsu unit in OpenRAN and 6G photonics push
- Keysight and NTT DoCoMo team up on 5G network emulation
Other articles on eeNews Europe
- First 300mm fab for the UK
- Patent war heats up as Samsung sued in Europe
- €3m for hybrid AI chip inspired by insects
- US venture firm launches custom AI chip business
- Nvidia opens up NVLink for chiplets to take on PCI Express
