What's New

Feature Article

Feature Article

Beyond the Qubit: Quantum Computing, Practical Alternatives, and Memory-Driven Computing

Hewlett-Packard Enterprise recently published a white paper (PDF, 770 KB) by Ray Beausoleil and Rebecca Lewington, arguing that for most applications of computing in the near future, novel approaches to classical computing may offer much greater performance than quantum computing.

They suggest that quantum computers may be ideal for modeling quantum systems, and for special problems such as decryption once noise problems can be overcome. However, for problems associated with Big Data analytics, a completely different approach is needed. The paper identifies an analog neuromorphic processor, the “Dot-Product Engine” built around a memristor array, as a more practical way to address the problems associated with Deep Learning on a large database.

They further identify a class of problems associated with optimization, such as the traveling salesman problem, which can be addressed by a novel Optical Computing technology, known as the Coherent Ising Engine. They project performance superior to that in superconducting quantum annealers, which have been proposed for similar problems.

Finally, they promote the paradigm of Memory-Driven Computing for Big-Data analytics, with close integration of memory chips and heterogeneous processors within a high-speed interconnect fabric.

A video addressing similar issues is also available on the HPE Discover website.

Technology Spotlight

Technology Spotlight

Delivering the Future of High-Performance Computing

Dr. Lisa Su, President of Advanced Micro Devices

At the recent DARPA Electronics Resurgence Initiative Summit, a keynote talk was given by Dr. Lisa Su of AMD, focusing on how the semiconductor industry is meeting the growing demands of future high-performance computing as Moore’s Law is slowing down.

Dr. Su explained that although Moore’s Law scaling to 5 nm is continuing, the pace of such progress is slowing down. But the performance improvement continues to develop at a rapid rate, due to a combination of three factors:

  • Microarchitecture on chip
  • Multi-chip packaging of chiplets
  • Integration of heterogeneous processors

Optimum system performance requires co-design of silicon chips, system architecture, and software. She presented the example of the exascale computer system being developed at Oak Ridge National Lab, which should exhibit 1.5 exaflops by 2021. This is a partnership of AMD and Cray, as is further described in this HPCwire article.

While the highest-performance chips and systems will initially be limited to the most expensive machines, it is expected that similar technology will become available within a few years in data centers, edge computers, and even mobile devices.

Watch the video presentation by Dr. Su.
Other videos from the Summit are available on the DARPAtv YouTube Channel.