Chicago Quantum Summit Observations (first of a series)

Originally posted here:
Observations from the Chicago Quantum Summit, and Quantum Engineering, The Next Technological Space Race, hosted by the University of Chicago, and organized by the Chicago Quantum Exchange (CQE)

• Event Date: November 8, 2018
• Publication Date: December 10, 2018
• Written by: Jeffrey Cohen, US Advanced Computing Infrastructures, Inc.
• Email: jeffcohenpersonal@yahoo.com, Cell: +1.312.515.7333

On November 8, 2018, I spent the day attending a U.S. Government, industry & academic conference hosted by the University of Chicago (via livestream). I attended the evening event, Quantum Engineering: The Next Technological Space Race, at The Standard Club in Chicago. The hosts were The University of Chicago, Argonne National Laboratory, Fermi National Accelerator Laboratory and the University of Illinois at Urbana-Champaign. There were nine guest speakers, from eight institutions, including The U.S. Congress, Google, IBM, Microsoft, the Office of Science in the U.S. Department of Energy, the U.S. Department of Defense, the U.S. National Science Foundation, the Physical Measurement Laboratory of the U.S. National Institute of Standards and Technology (NIST), with introductions from the four host organizations.
It was insightful and pragmatic. Different groups are shared their approaches, progress, requirements and views of the challenges to accelerate the advancement of Quantum Information Science and Quantum Computing. I left with an understanding of the ‘State of the Union’ on quantum computing and with a belief that Chicagoland is a hub of quantum innovation.
Key points I jotted down during the main event:

1. Teleportation of data. With a classical computer, a bit is either 1 or 0, and data is sent across the network as a stream of 1s and 0s to create a copy. With a quantum computer, a bit, or qubit, can be both 1 and 0 as a probability. A quantum network maintains entanglement between systems, which means they act as one system across locations. The data transmitted, if any in the classical sense, is teleported instantaneously (which could mean faster than the speed of light). A team of organizations, including the members of the Quantum Exchange, are building a quantum network testbed, about 30 miles long between Fermilab and Argonne, using existing dark fiber cabling, to connect their quantum systems.
2. Quantum Advantage. There is an upcoming point in the evolution of quantum computing when quantum computers can solve a problem significantly faster than ‘the fastest’ classical supercomputer. The type of problem could be factoring a large number, or a complex optimization problem. This point is called ‘Quantum Advantage’ as described by Harmut Neven, Engineering Director at Google Quantum Artificial Intelligence Lab. We are getting close to this point in time.
3. Throughput over number of qubits. There is a concept called Quantum Computational Throughput. This measures the amount of calculations that can be performed with a high degree of accuracy. Today’s quantum computers (think hardware and software stack) are relatively small and noisy, which is referred to as Noisy Intermediate Scale Quantum (NISQ) by Google. They have low but non-zero error rates for each calculation. Therefore, a series of calculations requires error correction and error mitigation to reach a reliable answer. The larger the system, the more error correction needed, with potentially no increase in throughput. We need to both advance the engineering of the stack to eliminate that error rate, while increasing qubit density (think more and better transistors on a chip).
4. Competing Commercial Approaches. Three industry players (IBM, Google and Microsoft) each described their own software stack, or framework, for their preferred quantum computing model. I just read that D-Wave has a fourth software stack for hybrid computing (quantum and classical).
5. Scarcity of talent. After improving the hardware, software and networking ‘stacks,’ the next challenge for quantum information science is scarcity of talent.  We heard there are ~20 million computer programmers. We know there are likely ~1 million physicists, mathematicians and materials scientists. However, there are only ~10,000 people able to write algorithms for quantum computers. The challenge is that performing ‘real research’ requires a combination of disciplines, most often in the same person.  The computer scientist needs to understand physics, or the physicist needs to learn computer science. This is rare to find. So, we either teach physics to computer scientists, or vice versa, to utilize the existing talent base. The sustainable, long-term approach is to start teaching multiple disciplines in the same program, and to younger (e.g., high school) students. 
6. Lack of sufficient supply chain. I recall hearing in multiple sessions that quantum hardware is in the experimental stage. Chips (containing qubits, gates, and connectors), are fully isolated and kept colder than deep space. Quality of raw materials, components and subassemblies must be very high, as failures can take a long time to diagnose and fix. There are few suppliers, if any, for most components, including the cooling material Helium-3. This limits the ability to rapidly scale production, and may require vertical integration of multiple hardware stack manufacturers.
7. Space Race Public Funding? If this is a type of ‘global space race,’ do we see significant U.S. Federal Government funding? This was asked in the public event. Prior public spending levels for individual programs were consistent, but relatively modest. The technology is too new to attract significant industry investment. However, spending may increase significantly with the passage of bill HR 6227, titled “ An Act to provide for a coordinated Federal program to accelerate quantum research and development for the economic and national security of the United States.” Its short name is the “National Quantum Initiative Act” and includes $255M per year for each of 5 fiscal years in spending by three coordinating Federal agencies.
8. Do we use quantum computing today? I believe so, for Atomic Clocks that are now accurate up to 1 second over 300 million years. The public key encryption foundation (which secures our public internet and most commercial communications) is now set to be replaced by 2025, due to the threat of quantum computers factoring large numbers fast enough to read encrypted messages before they lose their sensitivity. Other uses are from researchers, agencies, corporations, and individuals that use industry applications from D-Wave, IBM, Google and Microsoft in either pure, hybrid or emulator form.

Thank you for reading. Any errors, omissions, or opinions are my own. Please call or message me if you would like to discuss further.


In the second article I will cover recount the individual sessions. In the third article I will share my own conclusions, implications, and what it could mean for the general public.