Will real-time data processing replace batch processing?
At Confluent's user conference, Kafka co-creator Jay Kreps argued that stream processing would eventually supplant traditional methods of batch processing altogether.
Absolutely: Businesses operate in real-time and are looking to move their IT systems to real-time capabilities.
Eventually: Enterprises will adopt technology slowly, so batch processing will be around for several more years.
No way: Stream processing is a niche, and there will always be cases where batch processing is the only option.
Cloud Services / Edge Computing

ColdQuanta Delivers ‘Cold Atom Modality’ to the Cloud

ColdQuanta’s technique traps cesium atoms in a grid of laser light in a 2D array and then housing it in an ultra-high vacuum cell made by the company.
Sep 7th, 2021 8:00am by
Featued image for: ColdQuanta Delivers ‘Cold Atom Modality’ to the Cloud

The quantum computing space — much talked about over the past decade — is reaching a transition point where capabilities, demand and investment are pushing the market beyond the theoretical and ever closer to the commercial, according to Scott Faris.

The large traditional companies like IBM and Honeywell over the past few years have been joined by the likes of Google and Amazon and, more recently, a growing number of smaller pure-play vendors in driving the technology that is fueling the innovation. At the same time, the problems that quantum computing can address are becoming increasingly more significant and complex and now investors are making their presence felt.

“Some of the more traditional players in the marketplace are really starting to demonstrate that there is a path to build a product,” Farias, the newly named CEO of ColdQuanta, one of the emerging pure-plays, told The New Stack. “The engineering challenges are starting to be overcome. Largely, it’s been a combination of evolution of some of the parts and pieces necessary to make this work. It’s been an evolution of the science continuing to get a better understanding. The third component is it’s now a focus of capital. The first two were necessary, but that wouldn’t have worked unless companies were willing to commit the capital to go do the heavy lift. What’s been driving that is that the people were really getting quite nervous about what is next in terms of a great leap forward in computational capability.”

Faris is coming to ColdQuanta from Luminar Technologies, where he was chief business officer for a company that develops advanced sensor technologies for autonomous vehicles. In a number of ways, Luminar’s industry is similar to the quantum computing space of ColdQuanta, he said. Both have to rely on a lot of compute power, which has become a limiting factor as Moore’s Law has slowed.

“There’s a common roadblock where computing now is starting to be a factor in things not moving at the rate they have been,” he said. “Some people call this Moore’s Law really starting to show wear and tear. That’s really part of the story. The other part of the story is that the level of problems that we’re starting to face now are becoming so significant and so complex. That is really starting to drive the need. The amount of data that we’re generating to start thinking about solving problems is really starting to drive thinking about not getting incremental improvements of compute, but effectively getting exponential capabilities. All this stuff is just now starting to come together.”

A Growing Market

While quantum computing has been discussed since the 1980s, the market itself is still growing. However, companies and investors are starting to pour money into initiatives into a space that could lead to innovations in such areas as artificial intelligence (AI), pharmaceuticals, financial services and bioinformatics that couldn’t be accomplished with traditional compute systems. Market research firm Statista is predicting that the quantum computing space will grow to more than $1.76 billion in 2026; it will hit $472 million this year.

The smaller startups like ColdQuanta are looking to gain traction in a still-developing market dominated by the larger players, which along with those mentioned earlier include Microsoft and other lesser-known companies like D-Wave and Rigetti, which have been around for years. They also are leveraging a range of modalities — technologies for creating and managing qubits, or quantum bits — that include superconducting, ion trapping and photonics.

Cold Atom Modality

ColdQuanta and some others, such as Atom Computing, use the cold atom modality. ColdQuanta’s technique traps cesium atoms in a grid of laser light in a 2D array and then housing them in an ultra-high vacuum cell made by the company. The cell can fit in the palm of a hand; the computer itself is larger. Trapping the atoms with lasers enables ColdQuanta to cool them to a sub-5 microkelvin temperature, very near absolute zero. The lasers and microwave pulses are used to arrange them in a way to create the gates.

Atoms are smaller than qubits created by superconducting — a method used by IBM and Google, among others — so more can be brought together in smaller areas. Qubits are analogous to bits used in classical computing. However, while those bits a value of 0 or 1, qubits can be 0, 1 or both, creating the ability to run workloads and develop applications that are out of reach of classical computers.

Qubits also are notoriously fragile, so keeping them extremely cold is important to ensure they don’t flip from one state to another. ColdQuanta is able to keep the qubits at such low temperatures without the large liquid helium-filled dilution refrigerators that superconducting methods need.

‘Hilbert’ Makes Debut

The company took a significant step in July when it announced it had developed a gate-based 100-qubit computer called “Hilbert,” which will become available later this year, initially via the cloud on ColdQuanta’s platform. It could become available on other public clouds like Amazon Web Services, Microsoft Azure and Google Cloud in early 2022.

By comparison, IBM’s “Hummingbird” quantum computer has 65 qubits, though the company’s plans include a 127-qubit system by the end of this year and a 433-qubit computer in 2022. By the end of 2023, IBM expects to have a system with more than 1,000 qubits.

“What was exciting to me about ColdQuanta and others is that there’s been the traditional approaches where the big companies are focused, but now people are starting to realize, ‘OK, this is real, you can start to build product,’” Faris said. “It’s opened the door for ion trap or cold atom modalities to also attract investment and people to get behind some alternative pathways to accomplish this. In many ways, it’s where we’re trying to figure out how we go from the vacuum tube era or to the transistor era. … The forces at hand in multiple areas are starting to converge to say this is a real product, there’s a real solution here, and the demand is beginning to materialize.”

The Money Rolls In

Investments are coming into ColdQuanta. The company in May said it raised $20 million to bring on more people — it reportedly had 80 employees at the beginning of the year and plans to have 120 by the end of 2021 — and commercialize its technology and has raised a total of $74 million. ColdQuanta, founded in 2007, also was awarded more than $30 million in R&D funding from the U.S. government and a year ago received a $7.4 million contract from the Defense Advanced Research Projects Agency (DARPA) to develop a scalable, cold atom-based quantum computing hardware and software platform.

The issue of scalability — in terms of capability and capacity — is key to ColdQuanta and quantum computing, Faris said. The computational capability of a machine increases as the qubit count goes up and ColdQuanta can show a path where the count goes up quickly without major changes needed to the machine.

“Then there’s scalability in the other direction — there is a path that things get smaller and smaller and that unique combination of being able to drive computational intensity while at the same time reducing cost and footprint is something we’re all quite familiar with,” he said. “It’s what we call Moore’s Law. There is an argument made — at least within this modality — that there is a pathway to think about a Moore’s Law evolution of the technology and as things can get smaller, the computational power can go up and costs can go down over time.”

From Research to Commercial

ColdQuanta’s evolution is mirroring that of the quantum computing market as a whole, where it sits at the cusp of maturing from a research company to a commercial one. Along with announcing Faris’ appointment as CEO, company officials also created three business divisions to help scale and commercialize its products: Quantum Computing for developing the company’s cold atom platform, Quantum Research-as-a-Service (QRaaS) to create a research platform for developing prototypes, research and consulting services, and ColdAtom Technologies (CAT) for providing supply chain quantum devices and machines like glass vacuum cells and magneto-optical traps (MOTs).

Faris will be in charge of the rollout of the divisions. He has yet to take over the CEO job; he’s in a period of transition as he winds down his role with Luminar.

“There’s a full recognition of the scale of the challenge [at ColdQuanta], but I also argue the scale of the opportunity is commensurate,” he said. “This is ‘change the world’ kind of stuff, which is what I want to focus on doing.”

Group Created with Sketch.
TNS owner Insight Partners is an investor in: The New Stack.
THE NEW STACK UPDATE A newsletter digest of the week’s most important stories & analyses.