Unlike the historic microprocessor era, where computers were built with standard silicon parts, most computation is now performed by SoCs, where every new SoC design is an opportunity for application-specific customization. But realizing the potential benefits of greater specialization in commercial products requires advances in design productivity, as unsustainable SoC hardware and software design cost scaling is already a far greater challenge than transistor scaling. This productivity challenge lies behind the resurgence of interest in open-source hardware. The hope is that the hardware industry can replicate the success of the software industry, where innovative products are often created by mostly reusing existing code from open-source software stacks. While the open-source hardware movement has recently been driven by the widespread momentum behind RISC-V, a free and open ISA is only one component among many that will be needed to create a vibrant open SoC ecosystem. The architecture community is ideally placed to help fill out the open-source SoC stack and enable a new era of frictionless open-source hardware innovation.
Krste Asanovic is a Professor in the EECS Department at the University of California, Berkeley. He is also Chairman of the Board of the RISC-V Foundation, and is a co-founder and Chief Architect at SiFive. His main research areas are computer architecture, VLSI design, parallel programming, and operating system design, and he is co-director of the Berkeley ADEPT lab, which is aiming to reignite hardware innovation by reducing the cost and effort of deploying custom silicon for new cloud and edge applications. He is an ACM Fellow and an IEEE Fellow.
Increasing computing performance enables new applications and greater value from computing. With the end of Moore's Law and Dennard Scaling, continued performance scaling will come primarily from specialization. Specialized hardware engines can achieve performance and efficiency from 10x to 10,000x a CPU through specialization, parallelism, and optimized memory access. Graphics processing units are an ideal platform on which to build domain-specific accelerators. They provide very efficient, high performance communication and memory subsystems - which are needed by all domains. Specialization is provided via "cores", such as tensor cores that accelerate deep learning or ray-tracing cores that accelerate specific applications. This talk will describe some common characteristics of domain-specific accelerators via case studies.
Bill Dally is Chief Scientist and Senior Vice President of Research at NVIDIA Corporation and a Professor (Research) and former chair of Computer Science at Stanford University. Bill is currently working on developing hardware and software to accelerate demanding applications including machine learning, bioinformatics, and logical inference. He has a history of designing innovative and efficient experimental computing systems. While at Bell Labs Bill contributed to the BELLMAC32 microprocessor and designed the MARS hardware accelerator. At Caltech he designed the MOSSIM Simulation Engine and the Torus Routing Chip which pioneered wormhole routing and virtual-channel flow control. At the Massachusetts Institute of Technology his group built the J-Machine and the M-Machine, experimental parallel computer systems that pioneered the separation of mechanisms from programming models and demonstrated very low overhead synchronization and communication mechanisms. At Stanford University his group developed the Imagine processor, which introduced the concepts of stream processing and partitioned register organizations, the Merrimac supercomputer, which led to GPU computing, and the ELM low-power processor. Bill is a Member of the National Academy of Engineering, a Fellow of the IEEE, a Fellow of the ACM, and a Fellow of the American Academy of Arts and Sciences. He has received the ACM Eckert-Mauchly Award, the IEEE Seymour Cray Award, the ACM Maurice Wilkes award, the IEEE-CS Charles Babbage Award, and the IPSJ FUNAI Achievement Award. He currently leads projects on computer architecture, network architecture, circuit design, and programming systems. He has published over 250 papers in these areas, holds over 160 issued patents, and is an author of the textbooks, Digital Design: A Systems Approach, Digital Systems Engineering, and Principles and Practices of Interconnection Networks.
In 2015, US Chief Technology Officer Megan Smith raised profound questions about women's contributions in science, engineering and math being erased from history. In this talk we explore a case study of such erasure and surface a very counterintuitive conjecture about the underlying causes and effects.
For more information, please refer to "The Disappeared: Beyond Winning and Losing", which appeared in IEEE Computer in October 2018.
After earning her BS and MSEE from Columbia University, Lynn joined IBM Research in 1964, where she made foundational contributions to computer architecture. Fired by IBM as she underwent gender transition in 1968, Lynn started her career all over again in 'stealth mode'.
Joining Xerox Palo Alto Research Center in 1973, Lynn invented scalable MOS design rules and simplified methods for silicon chip design, was principal author of the famous 'Mead-Conway' text, and pioneered the teaching of these methods at MIT — launching a world-wide revolution in VLSI microelectronic system design in the late 1970s. Lynn also invented an ARPAnet based e-commerce infrastructure for rapid chip-prototyping in 1979, spawning the modern "fabless design" plus "silicon foundry" industry model for semiconductor design and manufacturing. Lynn joined the University of Michigan in 1985 as Professor of EECS and Associate Dean of Engineering, where she continued her distinguished career.
A Fellow of the IEEE and the AAAS, Lynn has won many awards for her contributions including the Computer Pioneer Award of the IEEE Computer Society, Wetherill Medal of the Franklin Institute, induction into the Computer History Museum's Hall of Fellows, election to the National Academy of Engineering, and four honorary doctorates. Awarded the 2015 James Clerk Maxwell Medal by the IEEE and the Royal Society of Edinburgh, her citation included these words: "Lynn Conway's work has provided the underpinnings for innovations, discoveries and achievements in every area of scientific and humanitarian study."