Work using cold atoms to leverage quantum effects for useful applications isn’t new. Think, for example, of atomic clocks and various sensing devices. But putting them to work as qubits in quantum computing is a relatively new idea. While ions have gained most of the attention here, one company, ColdQuanta, believes that using cold, neutral atoms (and Bose-Einstein condensate (BEC) effects) will end up being a much better approach that permits scaling to many thousands of qubits even while we are still in the so-called NISQ (noisy intermediate scale quantum) computing era.
DARPA seems to think there’s a fair chance that ColdQuanta is right and two weeks ago awarded the 13-year-old company a grant[I] to “develop a scalable, cold-atom-based quantum computing hardware and software platform that can demonstrate quantum advantage on real-world problems.” Currently, many qubit technologies are battling for sway – superconducting, optical, ion trap (also individual atoms, but charged), silicon spin, and a non-abelian anion. All of them have strengths and weaknesses.
“Over the next 40 months,” said Bo Ewald, ColdQuanta’s relatively new CEO. “We believe that we can scale this technology up and end up with a system that has thousands of qubits of pretty good fidelity, better connectivity, and more complicated gates than other approaches. By the end of the 40 months [length of the grant] we would be able to run this DoD real-world optimization problem as part of the program. It’s a radar coverage optimization problem.”
Ewald is a familiar figure in the quantum community. He was president of quantum pioneer D-Wave Systems for six years. While D-Wave’s quantum annealing technology approach has occasionally drawn mixed reviews, the company has nevertheless sold many machines, albeit research instruments, to industry and government. Ewald joined ColdQuanta roughly one year ago (March 2019).
Part of what makes the ColdQuanta so interesting is its breadth of aspiration. The company says its core technology has applications in everything from accelerometers, spoof-proof GPS devices, RF sensors, stealth communications, and much more. Quantum computing is just the most recent, and given the times, perhaps now the most visible initiative. In fact, ColdQuanta has been a supplier of a variety of quantum instruments and systems (including ion traps) for years. Two recent POC demonstrations of its cold atom (BEC) expertise went to the International Space Station – they include NASA/JPL’s Cold Atom Laboratory, CAL1 (2018) and CAL2 (2019), projects to test quantum matter behavior in space.
All qubit technologies have problems. One strength of using single atoms of the same element for a qubit is that by definition the qubits are identical. Also, neutral atoms, because they lack a charge, can be brought closer together more easily. The latter makes it easier to induce an interaction between outer shell (valence) electrons putting two atoms into “superposed state” to act as qubits. ColdQuanta leverages BEC characteristics and lasers to accomplish this.
Here’s a brief description from Ewald: “If you had a cloud of atoms of all of the same element, and you could cool them enough, very close to absolute zero, they would create a new form of matter. As you got them near absolute zero, they would coalesce into the Bose-Einstein condensate (some people describe this as a quantum gas). Steve Chu (MIT, later U.S. Secretary of Energy) showed in the 90s how you could use lasers to control atoms and won the Nobel Prize for it (physics, 1997). A laser has energy and mass and you can use it to hammer atoms and put tremendous force on the atoms, thousands of G’s of force on the atoms. In doing that, you can cause them to slow down or stop their motion. In your room, for example, atoms are moving around at around 600 miles an hour.”
“Researchers from the University of Colorado (CU) and from MIT followed up Chu’s work and demonstrated that if you created a highly-evacuated chamber and put atoms, thousands to a million atoms, of the same element in it, and then used lasers shining through windows in the chamber, you could stop the motion of the atoms enough that you would get them very near absolute zero. And bingo, they coalesce into this new quantum matter.”
Another CU researchers, Dana Anderson, went on to form ColdQuanta where he is now CTO. The original idea was to build laboratory equipment and systems to let people around the world build their own devices to create and manipulate quantum matter (BEC). This the company has done successfully.
“It turns out that if you don’t go quite as far as creating quantum matter (BEC), you can use the same basic idea and use a highly-evacuated glass, we’ll call it a cube, and put thousands to millions of atoms into it, and you can actually control those atoms, either as a cloud or on an individual basis,” said Ewald.
This is what ColdQuanta does. You need the right kind of element – one with favorable valence electrons. “You tend to use elements on the left side of the periodic table, so rubidium, for example, and cesium and others,” said Ewald. Under the right laser pressure, ‘lower energy’ wells are created, and it is possible to place atoms in the wells in an orderly configuration, and cause interaction (electron sharing) between atoms in neighboring wells. Currently, ColdQuanta is able to demonstrate at least 10×10 lattices with up to 100 qubits and the current technology should be possible to scale up to about 100×100 with up to 10,000 qubits. Ewald says next steps are to control a larger number of quantum logic gates.
This enhanced connectivity is a strength versus ion trap technology, where ions acting as qubits are typically aligned as a string of pearls. IonQ, an early pioneer using ion trap technology, has demonstrated lining up 13 ions of which 11 are functional. “I think there’s complete connectivity between those ions, but it’s a linear approach,” said Ewald who suggested that approach may be limited to 50 to 100 ions.
Both neutral atoms and ions boast longer coherence times than superconducting qubits but somewhat slower gate speeds. Also it’s worth mentioning that ion trap systems seem further along in development.
One of the cool things about the neutral atom approach is you don’t need the exotic refrigerators required by superconducting-based quantum computers. “That’s one of the great things about this technology. The lasers stop the motion of the atoms within this little glass cell, and so within the glass cell you’re very near absolute zero – you’re in the nano-Kelvin or micro-Kelvin range. But right outside the glass cell, it’s the ambient temperature. You don’t need a big cryostat. You know, you don’t need a dilution refrigerator. The cooling is really done by physics,” said Ewald.
Many challenges remain. For example, although the neutral atoms don’t repel each other as ions do they can drop out of position fairly easily. Said Ewald, “We have an atom tweezer that we’re working on. Suppose that we create this array of atoms and we’re missing one atom. We’re working on technology where you can effectively pick up an individual atom from the pool an insert it into the lattice if we’re missing one. So there’s a lot of work to be done, but my point is that our control over atoms is so fine that we can manipulate individual atoms.”
Ewald described the likely forthcoming ColdQuanta system as being comprised of a 3x3x3-foot cube housing the quantum part of the system – so fairly small – along with a few racks of traditional systems and electronics to control the lasers, etc. He estimated the system will require on the order of 10Kv.
A quantum computer BEC-based machine, of course, has still not been released; Ewald suggested one will be relatively soon with broader availability via the web perhaps in the fall.
The company, not surprisingly, will build its own low-level programming tools and environment, but emphasizes the plan is to work with the rapidly growing open source and commercial quantum software ecosystem. Access to the systems will likely include web-based access and perhaps sales of systems to customers.
It is clearly still early days for ColdQuanta’s quantum computing efforts.
As for near-term milestones, Ewald said “The two biggest visible milestones so far were the two launches of our technology on to space station. That proved the ruggedness of our technology. On the quantum computing side, the DARPA award is very significant for us. I think over the course of the next year or so, we should be able to demonstrate the technology at least starting to approach a hundred or so qubits that function and that you can do something with. Then over the next three years as part of the program, we [expect] to be getting into the thousands of qubits.”
Link to Ewald presentation: https://www.youtube.com/watch?v=OrEb0SQHdbY&feature=emb_rel_end
[i]Others on the grant include defense contractor Raytheon Technologies, Wisconsin–Madison, Argonne National Laboratory, University of Chicago, NIST Gaithersburg, University of Colorado Boulder, University of Innsbruck, and Tufts University.