D-Wave Systems today launched its cloud platform for quantum computing – Leap – which combines a development environment, community features, and “real-time” access to a D-Wave 2000Q quantum computer including one minute of free runtime or enough time to run 400-4,000 experiments according to D-Wave. This is the second such announcement in a month. Rigetti Computing launched its Quantum Cloud Services (QCS) in early September. Longtime quantum player IBM launched its cloud offering, the IBM Q Experience, back in 2016.
The idea all three companies share is to leverage cloud delivery of training and quantum compute time to accelerate development of a quantum computing ecosystem, particularly among developers and users unfamiliar with quantum computing.
“We’d like to sort of unlock the power of quantum computing for potentially hundreds of thousands of developers who have heard about quantum computing and they want to move in the direction to build and run their own quantum applications,” said Murray Thom, D-Wave director of software and cloud services, in a pre-launch briefing with HPCwire that included access to the Leap platform. Thom is a 16-year veteran of D-Wave, which itself is coming up on 20 years (founded in 1999). “My quantum experience is now old enough to drive,” he jested.
To Thom’s latter point, a lingering question for quantum computing is not whether it is old enough but whether it is (nearly) ready enough to leave the garage and drive onto main roadways. The cloud offerings make taking quantum computing test drives much easier.
Of the varying approaches to quantum computing, D-Wave’s quantum annealing approach is among the furthest along. The company was founded in 1999 and notes on its website, “[D-Wave] systems are being used by world-class organizations and institutions including Lockheed Martin, Google, NASA, USC, USRA, Los Alamos National Laboratory, Oak Ridge National Laboratory, Volkswagen, and many others. D-Wave has been granted over 160 U.S. patents and has published over 100 peer-reviewed papers in leading scientific journals.”
IBM, the granddaddy of the quantum pack with related research stretching back to the 70s, says its “cloud IBM Q Experience has more than 97,000 users who have run more than 6 million experiments. And more than 120 research papers have been published based on experiments run on these systems. The IBM Q Experience is also part of the curriculum at more than 1,500 universities, 300 high schools, and 300 private institutions.”
While those market traction and research activity numbers are impressive, in practical terms quantum computer capacity (number of qubits) and reliability remain problematic. Currently, IBM Q offers 5-qubit and 16-qubit processors. Rigetti offers a 16-qubit processor with plans to scale it up to 128 qubits within roughly a year. There is broad agreement in the quantum computing community that many more qubits are needed to tackle practical applications.
The D-Wave 2000Q used by Leap is a 2,000-qubit machine and significantly larger than those of its rivals. That said, D-Wave’s quantum annealing technology is very different from gate-based models (universal quantum computers.) Within the constraints of its quantum annealing approach, D-Wave systems are very well-suited for solving problems in optimization, machine learning, and materials sciences says the company. It’s been racing to stimulate development of applications.
So far, none of the quantum computing suppliers or their pioneering users have demonstrated so-called “quantum advantage” – applications in which quantum computers are distinctly better than classical computers. But that day is coming, perhaps soon, they all believe. Rigetti is even offering a $1 million prize for the first to do so on its QCS.
D-Wave continued to beat the application progress drum in today’s Leap announcement, “To date, D-Wave customers have developed 100 early applications for problems spanning airline scheduling, election modeling, quantum chemistry simulation, automotive design, preventative healthcare, logistics and more. Many have also developed software tools that make it easier to develop new applications. These existing applications, tools, and community give developers a wealth of examples to learn from and build upon.”
In what Thom called a ‘happy coincidence’ Leap is being launched during a D-Wave user group meeting in Knoxville, TN. “This is our fourth user group meeting. We started in 2016. Last year there was one in Washington, DC. There are on the order of 80 customers, developers, and researchers who come to talk about tools and methods for programming and applications,” he said.
Here are a few Leap highlights:
- Free access: free, real-time access to a D-Wave 2000Q quantum computer to submit and run applications, receiving solutions in seconds.
- Familiar software: the open-source Ocean software development kit (SDK), available on GitHub and in Leap, has built-in templates for algorithms, as well as the ability to develop new code with the familiar programming language Python.
- Hands-on coding: interactive examples in the form of Jupyter notebooks with live code, equations, visualizations and narrative text to jumpstart quantum application development.
- Learning resources: comprehensive live demos and educational resources to help. developers get up to speed quickly on how to write applications for a quantum computer
- Community support: community and technical forums to enable easy developer collaboration.
“Even at launch the system is prepared to handle tens of thousands of users,” said Thom. “We have a quantum computing system, which is the primary online system available for this, and we also have a secondary quantum computing system available as backup. The Leap front end is hosted on Amazon but it’s built to run in any public cloud. The users will not be charged nor do they need accounts.”
HPCwire’s brief session on Leap suggested it’s an easy-to-navigate platform rich in resources for relative novices and those more experienced with quantum computing concepts. Attracting users steeped in traditional HPC and cluster computing paradigms and enabling them to engage in quantum computing without getting bogged down in quantum theory is a major goal. It’s one of the reasons, for example, D-Wave’s tools are Python-based.
Thom said, “Right up front [on the Leap dashboard] we made available information where they can learn about Leap, about case studies, about quantum application development, also information and tutorials about quantum computing, a tour of our laboratory, and some really interesting videos that our customers have put up about their projects.” The price for free access to Leap, said Thom, was a requirement that all software developed on the cloud platform be put into open source. Software developed offline but run online need not be. Users can upgrade from their allotted one minute of free time to paid time starting at $2,000/hr.
With three quantum cloud platforms now available, it will be interesting to see whether development efforts do indeed accelerate and begin to generate new applications. One can imagine researchers kicking the tires on all three platforms.
Thom said, “I’ve already begun to see that in terms of researchers who have been looking to get access to multiple systems trying to see if they can find problem instances they can run on multiple platforms. I don’t suspect that will be every user. I think that people will gravitate towards opportunities to learn more about the system and opportunities to leverage the platforms to actually get work done. I think they will probably very quickly gravitate to those types of environments well suited for their applications.”
Leap will initially be a development platform – no one is truly using quantum computing yet for ‘production’ purposes – but it was designed to support production requirements. “We anticipated there will always be some customers who are interested in or required to have their own systems in-house and other customers who are interested in basically having their systems hooked in remotely and being able to access that way. Having said that, I anticipate the platform will evolve as the community is developing applications and we start to learn more about their use cases and their workflows and data and we’ll evolve this system to adapt to that,” said Thom.
Quantum industry watcher Bob Sorensen, VP of Research and Technology, Hyperion Research, said Leap is an important step in the right direction. “I see D-Wave’s new software infrastructure as a significant advancement in moving quantum computing away from the realm of being seen as mysterious to where software developers don’t have to fully appreciate the underlying quantum physics to work on QC algorithms and applications in a more familiar traditional programming environment. This can only encourage the proliferation of QC software developers and compelling QC-based use cases.”
The proof will be in the payoff. Thom noted two milestones.
“When the growth of the community reaches a critical size there’s likely to be an emergence of sophisticated, shared open source software being used for programming these systems. More than likely, people within that community will start to making demonstrations where they can say in this particularly application domain I can do better than industry. That would be a very significant milestone. There will also be a milestone where someone discovers the equivalent of a quantum killer app. That’s going to lead to a demonstration of quantum advantage with the application. The progression of the quantum industry will rapidly change at that point.”
Link to D-Wave announcement: https://www.dwavesys.com/press-releases/d-wave-launches-leap-first-real-time-quantum-application-environment