Quantum computing has emerged as one of the most talked about emerging technologies in Silicon Valley and beyond–but it is also perhaps one of the most poorly understood. The most pressing questions about quantum computing come from the enterprise user community, which is searching for clarity on the following points: What types of real-world problems can quantum computers (QC) solve, and how can enterprises begin the application discovery process?
NextBigFuture met with QC Ware’s CEO (Matt Johnson) and Head of Business Development (Yianni Gamvros) to find answers to these questions.
Introduction to QC Ware
QC Ware is a quantum computing software company based in Palo Alto, and it is now opening its European headquarters in Paris. The fast-growing firm has established itself as a leader in the nascent QC software space and is working with a number of Fortune 100 companies to develop prototype solutions and algorithms for some of their hard computing problems.
QC Ware has developed a hardware-agnostic software platform that serves as the “delivery vehicle” for those applications. The platform also provides a much-needed abstraction layer, which allows classically-trained data scientists to run problems on prevalent QC hardware stacks that are currently commercially exposed.
QC Ware’s Mission
Matt explained that QC Ware’s mission is to be a leading software solutions provider for enterprises, and that the company is playing the long game to achieve that goal. QC Ware has assembled one of the largest teams of quantum algorithms researchers in industry, and their sole focus is exploring methods to squeeze power out of the early-generation quantum processors that are being fielded by QC hardware developers such as IBM, Google, Rigetti and D-Wave Systems. The quantum computers that are being fielded today are referred to as having “noisy” processors, which means they are not fault-tolerant and hence calculations are subject to error.
Quantum Algorithm Experts
And that is where QC Ware’s team of algorithm experts come into play. They work to express these problems in ways that make it possible to execute them on noisy quantum processors. Their goal is to construct so-called shallow-depth circuits on which the problems can be run. For the non-computer scientists in the room, that means they find mathematical shortcuts to minimize the number of steps (e.g., gate operations) required to solve the problem on a quantum processor.
Solutions for Practical Problems and Applications
So what kinds of practical problems can these computers solve? Broadly speaking, quantum computers are able to solve problems in optimization, simulation, and machine learning. QC Ware is already running small-scale versions of such problems on quantum computers, but there are two things preventing QC systems from outperforming today’s “classical” computers. The first is the sheer size of today’s quantum processors. They lack enough physical qubits (think quantum transistors) and sufficiently high-quality qubits to produce any “wow” factor that an ROI-sensitive buyer or user cares about. In addition, the few useful quantum algorithms that have been invented to date (those that have a provable speedup over classical algorithms) need to be extended to handle the idiosyncrasies of practical computing problems. QC Ware is leaving the first problem in the hands of the QC hardware developers and is focusing exclusively on the algorithm/application development challenge.
QC Ware is employing this expertise with companies in various sectors including financial services, oil/gas, aerospace, and automotive. These companies all have computing bottlenecks, and they would have a competitive advantage if they were able to solve those problems faster.
Value to QC Ware Customers
Yianni explained that the immediate benefit to getting involved in quantum computing today for enterprises is to start building the necessary skills to take advantage of quantum computing as soon as it is ready for production problems. Whoever starts investing now will be better prepared for the future when quantum computing will disrupt existing industries and create new ones in the same way machine learning and AI is doing today.
QC Ware’s objective while working with large companies is to build a software bridge between the API’s that the QC hardware vendors provide and the problems that the enterprise user has. QC Ware’s cloud platform provides users with a hosted jupyter notebook that has QC Ware’s python libraries pre-loaded and -configured. Users can also get an API key and use the python API from local installations. The company will also roll out APIs in other high-level languages such as Java and C++.
One unique attribute of the service is that it allows users to run problems on a variety of QC hardware platforms. QC Ware has managed to abstract away the programming complexity for each of the QC hardware platforms currently exposed so that users can easily run a particular job on any platform without having to re-formulate the problem or undertake any additional programming steps.
The company tested an alpha version of its service since last summer, and it will be launching an expanded beta version later this year. This version of the service will be made available to commercial enterprise users and government agencies–either as a stand-alone service or as part of a software development project.
Although commercial QC deployment is several years away, QC Ware has developed a successful business that allows it to grow prior to the hardware “getting there”: The company generates revenue from developing prototype applications and from selling subscriptions to its cloud service. The cash generated from those services is plowed back into what QC Ware calls applied research. Its algorithm team spends its time figuring out how to map enterprise problems onto quantum algorithms, and to run those problems on QC hardware platforms. The work product of the research is function libraries and specialized algorithms that make their way into QC Ware’s cloud service.
Viewed from that angle, QC Ware’s business is on its way to becoming a force that is equal parts Oracle and Redhat. Oracle, in that it is making a hard-to-use resource accessible to the broad enterprise community. Remember that back in the 1970’s, relational databases were viewed as unbuildable curiosities and would be impossible to manipulate. Redhat, in that QC Ware’s software platform, sits on top of the basic API’s provided by QC hardware vendors. Their platform encapsulates an interface that allows classically-trained enterprise users (with no presumed QC knowledge) to interact with a stable, commercial-grade cloud service.
Both Matt and Yianni have a realistic view of the technical challenges that must be addressed to make QC a useful resource for enterprises. They are nonetheless moving out at a fast pace on software development because they are taking the view that the QC hardware vendors will solve those engineering and materials problems.
There is some risk attached to QC Ware’s thesis, but then again risk is not a foreign concept in QC Ware’s hometown of Palo Alto.
SOURCES- Interviews, QC Ware
Written By Brian Wang, Nextbigfuture.com