Technology Spotlight - 2017
Videos from the IEEE Industry Summit on the Future of Computing
The 1st IEEE Industry Summit on the Future of Computing was held on November 10, 2017 in Washington, DC, as part of IEEE Rebooting Computing Week. Many of the presentations of the Summit were recorded by IEEE.tv, and will be available for viewing in the near future. Two presentations are already available:
- Dr. Dario Gil, IBM Research, Yorktown Heights, New York
Dr. Gil presented an overview of futuristic research projects at IBM, and their implications for the future of computing.
- Prof. Bing Liu, University of Illinois at Chicago
- Dr. Jeronimo Castrillon, Center for Advancing Electronics Dresden (CfAED), Germany
- Dr. Robert Voigt, Northrop Grumman Corp., Baltimore, Maryland
- Dr. Dario Gil, IBM
This included brief comments by the panelists on a variety of future computing technologies, followed by questions and discussion from the audience.
Check again at IEEE.tv for additional presentations from the Industry Summit as they become available.
The 3rd annual LPIRC competition was held in Honolulu, Hawaii, July 21, 2017, co-located with the Conference on Computer Vision and Patterns Recognition (CVPR). This was sponsored by IEEE Rebooting Computing, together with the Green ICT Initiative. Dramatic improvements in low-power image recognition are necessary for mobile devices with limited batteries. Contestants were required to recognize thousands of images within just a few minutes, using computing resources with very low power consumption. The winning team of students from Seoul National University in South Korea used only 2 watt-hours of energy to recognize 20,000 images. This was several times better than the winners in previous competitions.
For further information on the LPIRC 2017, see here.
For a video from IEEE.tv on LPIRC 2017, see here.
Plans are underway for LPIRC 2018 to be held with CVPR in Salt Lake City, Utah in June 2018.
Keep watching for the announcement!
At the MIT Venture Capital and Innovation Conference in Feb. 2017, Dr. Dario Gil, VP of IBM Research, presented a simple overview of quantum computing and what kinds of problems it may be useful to solve. The key is that a quantum computer with N coupled quantum bits (qubits) has equivalent computing power to a classical computer with 2N parallel bits. For a quantum computer with large N (which does not yet exist), this exponential speedup could address a range of computational problems that are currently unsolvable, including decryption, quantum chemistry, and combinatorial optimization. Dr. Gil also presented recent IBM research on development of quantum hardware and software.
See here for the video of Dr. Gil’s presentation.
For further information on the MIT VC Conference, see here.
At the International Solid State Circuits Conference in San Francisco in February, Dr. Ahmad Bahai, CTO of Texas Instruments, argued that exponential growth will be maintained in the performance of a variety of electronic and computing applications, even as the traditional Moore’s Law scaling of transistor size and density stalls. This will require engineering innovations including 3D integration, novel circuit and system architectures, and improved packaging and manufacturing.
Other plenary talks from the 2017 ISSCC are available here.
At ARM TechCon 2016 in Santa Clara, CA, Dr. Greg Yeric, ARM Fellow and Director of Future Silicon Technology, reviewed the current status of Moore’s Law, and the efforts to scale transistor gate length down to 5 nm and 3 nm. While others have emphasized how Moore’s Law is ending, Dr. Yeric focused on continuing improvement for at least the next decade. This will require not only improved design and fabrication tools, but also new materials and devices to be integrated into silicon technology.
Slides from Dr. Yeric’s presentation are available here (PDF, 3 MB).
HPE sponsored Discover 2017 in Las Vegas, Nevada, June 5-8, 2017, exploring some novel technologies for the future of computing. This video includes presentations by Kirk Bresniker, John Paul Strachan, and Thomas Vaerenbergh on memory-driven digital computing, memristor-based analog computing, and silicon photonic computing.
Other videos from the same conference are available here.
In May 2016, IEEE announced the formation of the IRDS™ as part of the Industry Connections Program of the IEEE Standards Association, under the sponsorship of IEEE Rebooting Computing. In March 2017, the IRDS™ team published nine foundational white papers that outline the technical components required to create an official roadmap in early 2018. To help us better understand the IRDS™ program and the roadmap it will soon produce, we interviewed three participating experts: Erik DeBenedictis and Matthew Marinella of Sandia National Lab and Geoffrey Burr of IBM. To listen to or download this or other RC-related podcasts, see here.
IRDS™ will have a preliminary release of their 2018 roadmap in November 2017 as part of Rebooting Computing Week.
The South-by-Southwest Conference in Austin TX, Nov. 10-19, 2017, included a conference track on “Intelligent Future”, with a wide range of sessions related to artificial intelligence and the future of technology.
This, in turn, included an IEEE-led panel discussion on “Going Beyond Moore’s Law”, featuring IEEE Rebooting Computing Co-Chair Tom Conte. This video from IEEE.tv provides clips of the panel presentation and interviews with some technology leaders in attendance.
The Supercomputing 2016 Conference (SC16) was held in Salt Lake City, UT, Nov. 13-18, 2017.
One of the invited talks was by Prof. Thomas Theis of Columbia University. He showed that while existing device technologies may be sufficient to achieve exascale performance, further increases in performance will require new devices and architectures. These may include low-energy approaches such as tunneling FETs, spintronic devices, and neural networks.
Dr. Theis also co-authored an article on a similar topic in the latest issue of Computing in Science and Engineering, available here.
Videos of the other invited talks from SC16 are available here.
The 2016 ICRC was held in San Diego, Oct. 17-19, 2016. As part of a government funding panel, Dr. William Vanderlinde of the US Intelligence Advanced Research Projects Activity (https://www.iarpa.gov/) discussed IARPA research programs in support of the National Strategic Computing Initiative. Specific projects include such unconventional computing approaches as quantum computing, superconducting computing, and neuromorphic computing.
The 2016 ICRC was held in San Diego, Oct. 17-19, 2016. David Mountain of the Laboratory for Physical Sciences in Maryland presented a talk comparing digital CMOS to analog memristor implementations of neural networks for scaling to neuromorphic computers.
The 2016 TTM meeting sponsored by IEEE Future Directions was held in San Diego, Sept. 20-21, 2016. The theme was “Making the Future”. Many of the presentations were recorded by IEEE.tv. Watch the presentations here.
One presentation was a panel on Rebooting Computing, with panelists Elie Track, Tom Conte, Stan Williams, and Bob Voigt.