After decades as an emerging technology, quantum computing is moving rapidly toward commercialization. And in doing so, it is increasingly coming full circle back to Boulder, where some of the earliest breakthroughs were made.
IonQ, the largest independent player in quantum computing, has chosen Boulder as the site of a new research and development facility, joining Google Quantum AI, which made a similar move in March.
“This is a big bet for us,” said IonQ chairman and CEO Niccolo de Masi, during a ceremony on Tuesday to unveil a 22,000 square feet lab at Boulder 38, which is at the intersection of 38th Street and Arapahoe Avenue.
The company plans to invest $100 million in the new R&D facility, which is expected to be completed in late summer to accommodate new quantum computers arriving from the United Kingdom later this year.
Once those are set up, a team of engineers and computer scientists will monitor hardware performance and test how they handle different software configurations and algorithms. The location will also serve as a quantum data center handling real-world applications.
In February, the Colorado Economic Development Council approved $2.76 million in Job Growth Incentive Credits for IonQ, conditioned on the creation of up to 150 new jobs paying an average annual wage of $168,422.
The EDC has also designated those state tax credits as refundable, meaning they could be sold to raise funds. Boulder’s Gunbarrel area also received a CHIPS Zone designation in November, which opens up additional credits.
Incentives aside, Boulder beat out its competitors because of the technical expertise residing at the University of Colorado and the National Institute of Standards and Technology and other federal labs, IonQ executives said.
The Boulder area has the highest concentration of quantum researchers, companies and workers in the world. Getting those experts to leave is a tough task, said David Allcock, vice president of Quantum Computing Science with IonQ, and the executive who will be in charge of the new lab.
Although IonQ is based in Maryland, the company’s core technology, like so much in quantum computing, traces back to Boulder.
Company co-founder Chris Monroe, together with Nobel-prize-winning physicist David Wineland, developed the world’s “quantum logic gate” in 1995 while working at NIST, home to the atomic clock.
That set the stage for the use of trapped ion technology in computing, which was further developed at the University of Maryland.
“In so many ways, we see IonQ’s decision to choose Boulder as the site for this facility as a reflection of the community’s longtime commitment to innovation, entrepreneurship and emerging industries,” Boulder Mayor Aaron Brockett said.
IonQ went public in November 2020 and on Thursday had a market value of $21.4 billion, making it the largest standalone quantum computing company in the world.
Its growing financial cloud has allowed it to make several big acquisitions, including the $1.8 billion purchase announced earlier this year of SkyWater Technologies, a large U.S. semiconductor foundry based in Minnesota.
In January, the company ., a California company with a large presence in the Broomfield area. Skyloom has developed technologies for transmitting data via lasers through space instead of through land-based fiber optic cables.
“The conversation is shifting from can we do quantum to how do we scale it,” said Jessi Olsen, CEO of Elevate Quantum, a Boulder-based association working to promote the commercialization of quantum computing within the state. “Colorado is where the quantum future is being built.”
Google Quantum AI, a subsidiary of Alphabet, is another leading player in the quantum space that set up shop in Boulder after previously concentrating most of its investments in California and Washington.
In March, it established a new research and development team under the direction of CU Boulder professor Adam Kaufman to build quantum computers using what are known as neutral atoms.
The , which has primarily focused on superconducting or the use of extremely cold temperatures to harness atoms.
Practicality versus supremacy
Quantum computers operate at a power of magnitude so far beyond existing systems that it can be hard to describe them without evoking terms from the supernatural.
Classical computing systems are binary, where bits function like light switches, either on or off. Information is processed in a linear sequence, one step at a time. The bits can be in one place at a given time, and their paths can be traced, making error detection and correction possible.
Quantum computers use “qubits,” which can access subatomic states. One state is , allowing them to be both on and off at the same time and in multiple places at once. And there is entanglement, where qubits can be linked regardless of distance, say a computer in Boulder and one in Maryland, using a Skyloom laser.
If bits are mere mortals, occupying space and time, qubits possess omnipresence and telepathy. But they are “wild” in their natural state and not easy to tame.
Historically, extremely cold temperatures, near absolute zero, were used to put them in a quiet state, which required bulky and expensive systems. IonQ, by contrast, initially used a series of laser “lassos” to control the atoms and drain them of their kinetic energy.
While a big step forward from cryogenic systems, scaling requires a complex array of precise lasers. IonQ will be testing systems in Boulder that use on-chip electronics to do more of the gatekeeping, making for an easier commercial transition.
Quantum computers can handle a massive number of mathematical possibilities at once and complete calculations once considered impossible. Computations that the world’s most powerful conventional supercomputers could never hope to complete can be done in minutes and hours with quantum computers.
But qubits, despite their superpowers, are incredibly fragile and require isolation from the outside world.
Even the tiniest disruptions, whether in temperature or electromagnetic waves, can make them unstable. Qubits are so unstable that even observing them to see if they are behaving properly can upend the process.
For years, adding more qubits to correct errors was the attempted solution, but it only increased unreliability. Recent advances have resulted in lower error rates as more qubits are added.
IonQ uses individual atoms of Ytterbium and Barium as qubits within an ultra-high vacuum, protecting them from the outside world. Once trapped, the ions are reduced to a ground state through lasers, similar to what freezing them at absolute zero achieves.
Rather than focusing, as Google has, on “quantum supremacy,” IonQ has put its energy into running a smaller number of qubits at a much higher accuracy rate. All major cloud providers, including AWS, Azure, and even Google Cloud, are customers.
Its trapped ion systems are shrinking in size and can operate at room temperature with fewer lasers; they can now fit into rack-mounted systems compatible with standard data centers.
One rack-mounted IonQ system could, in theory, replace thousands of energy-hungry GPUs for AI training tasks, said Chris Ballance, president of Quantum Computing at IonQ.
Investors are pouring billions of dollars into AI-related companies that are promising to resolve bottlenecks, from building more data centers to providing more power to increasing chip capacity to photonics.
But quantum computers have shown they can learn complex patterns from much smaller datasets than classical AI and they can resolve complex problems using a fraction of the energy.
If conventional chips represent the information age equivalent of fossil fuels, quantum chips represent the coming “clean energy” alternative.
“It can solve problems for far less power,” Ballance said.
That in turn could make the expansion of AI, which relies heavily on data centers, more palatable to the public.
Gallup, , found that 7 in 10 Americans are opposed to the construction of data centers in their communities, with nearly half “strongly opposed.”
Key concerns include the raw land that centers use, the diversion of water resources to cool servers and electricity requirements, which is putting strains on the grid and driving up residential electricity prices.
Americans are now more accepting of nuclear energy plants in their communities than data centers, Gallup has found.
Because quantum computers can handle complex calculations in a fraction of the time conventional computers require, less energy per solution is required.
One transition scenario involves hybrid data centers where quantum machines do the heavy mathematical lifting while traditional servers handle the interactions with users.
How fast that day arrives remains to be determined. But the new research facilities that Google and IonQ are launching in Boulder could play a key part in whatever comes next.
