In 2018, Atom Computing used its initial seed funding of $5M to secretly begin building a proof-of-concept quantum platform equipped with 100 qubits. On the strength of its technology and increased investor confidence, Atom took in another $15M in Series A funding in mid-2021 plus three grants from the National Science Foundation. The Series A funds were used to build a larger team of quantum specialists. Earlier this year, the company closed another $60M in Series B funding. Those funds will be directed toward building a much larger second-generation quantum computing system that can run commercial use-cases.
Following a number of accomplishments in 2021, last week Atom Computing announced it had set a qubit coherence time record that was longer than any other commercial quantum platform.
Insider Podcast with Rob Hayes, Atom Computing CEO and Patrick Moorhead, MI&S CEO and Chief Analyst, discuss Atom Computing and the state of quantum computing.
It is discussed later, but a quantum computer is largely defined by its qubit type. Qubits (quantum bits) are the basic unit of quantum information in quantum computers, and coherence expresses the length of time a qubit can maintain its quantum state. Atom Computing creates spin qubits by using unique energy levels in the isotope Strontium-87.
I had the opportunity to discuss Atom Computing’s qubit coherence and other features with Dr. Mickey McDonald, Senior Quantum Engineer for Atom Computing. Dr. McDonald was awarded a Ph.D. in Atomic and Molecular Physics from Columbia University as well as a Masters from Columbia. He received his BA in Physics from Cornell University.
How long is 40 seconds?
Atom Computing’s record-setting qubit coherence was 40 seconds plus or minus 7 seconds. That may not seem very long compared to everyday life, but relative to quantum states, it is more than several lifetimes. Coherence times vary widely from a few milliseconds and up, depending on qubit type, hardware configurations, and operational procedures.
I asked Dr. McDonald about the natural coherence of Strontium-87 qubits. He said, “When a strontium atom is prepared in one of its nuclear spin states, it can coherently remain in that state for a very, very long time. We engineer two of those nuclear spin states to serve as the zero and one states of our qubit.”
Coherence is vital because qubits must maintain a quantum state long enough to run a quantum circuit. Imagine trying to add a long list of five-digit numbers on a calculator with a weak battery only to have it go blank halfway through the list. That is what short coherence feels like with a deep quantum circuit.
Long coherence allows a computer to run deeper quantum circuits and aids in quantum error correction research by providing more time to detect and correct errors without completely degrading purposeful computation.
There are various types of qubits in use. Today’s prototype quantum computers use superconducting qubits (IBM) and trapped-ion qubits (Quantinuum and IonQ). Other qubit technologies include quantum dots, NV center, electron on helium, photons, and topological.
All qubits are fragile and susceptible to errors caused by interactions with environmental factors such as electrical noise from internal components, background radiation, other qubits, wiring, and much more. These interactions can cause a qubit’s quantum state to collapse which creates errors and an irretrievable loss of information.
Atom Computing uses nuclear-spin qubits created from a neutral atom of Strontium-87 that stores information in its nuclear spin states. Neutral atoms have been studied for decades and in fact, a proposal to use neutral atoms for quantum computing was proposed more than 20 years ago. It was not until recently that the necessary technologies have been able to be integrated into a high performance quantum computing system.
Also of interest is that some of the supporting technologies needed for neutral atoms, such as optical tweezers and laser cooling, were recipients of Nobel prizes.
Dr. McDonald explained the nature of the qubit’s accuracy. “Strontium-87 may sound esoteric, but it has been used for almost two decades to create extremely precise atomic clocks.”
The technology offers other advantages as well:
- Natural qubits are perfect. Every qubit is identical to all others of the same species
- High connectivity among qubits
- Scalable to large numbers of qubits
- Qubit arrays offer flexibility as 2D or 3 D
- Wireless control using lasers
- Optical tweezers are scalable, allowing for easy mobility
- Proven science over decades of physics research – this atom type is used in atomic clocks because of its accuracy
Creating neutral atom spin qubits
Creating neutral atom spin qubits may sound like a complex process, but that is because it is a complex process.
First, a stream of hot strontium atoms flows into the vacuum chamber. Next, multiple lasers bombard the strontium atoms with photons and magnetic fields that slow atom momentum to a near motionless state, lowering the atom’s temperatures to about one-millionth of a degree above absolute zero. Lasers eliminate the need for cryogenics, making it easier to scale qubits.
Optical tweezers are fabricated in a glass vacuum chamber and used to trap atoms in a lattice of interlacing laser beams.
Manipulating a nuclear spin qubit
Atom Computing has chosen to use strontium-87 spin qubits for several reasons:
- This isotope provides nuclear spin levels which can be leveraged as qubit states.
- These qubit states are largely insensitive to external sources of noise, thus decreasing errors and improving coherence times.
Strontium-87 has an atomic structure with ten nuclear spin levels. However, a spin qubit only needs a level to represent the quantum state of 1 and a level to represent the quantum state of 0. Thanks to the nature of neutral atoms, a laser can push the unwanted eight levels off to the side, leaving two desired levels intact and available for manipulation.
Creating a superposition
Two electronic states defined by the nuclear spin of the strontium-87 atom represent the zero and one states of the qubit. Moving from one state to the other requires an operation called a two-photon transition which is accomplished by simultaneous application of two lasers that shift the qubit from one spin state to the other. The length of time the lasers are applied, as well as their relative phases, control the resulting superposition state of the qubit. Driving a coherent operation for both ground states, place the qubit in a superposition of 1 and 0, readying it for use.
“Basically,” Dr. McDonald said, “we work hard with strontium 87 to turn a naturally occurring ten-level system into an engineered two-level system. We can then manipulate that two-level system by applying two-photon transitions to the atom.”
Arbitrary operations on qubits in parallel
The researchers at Atom Computing also developed an equally important quantum capability, which was part of the published research paper but largely ignored by the media. The development consisted of a software-controlled laser scheme that performs arbitrary operations on multiple or individual qubits in a column or a row. Atom Computing demonstrated this with a 21-qubit array consisting of 3 columns of 7 qubits each.
Dr. McDonald said, Dr. McDonald said, “I referred to the system as Software Configurable Dynamic Lasers because the software allows you to control the phase and amplitude of individual lasers and then imprint that information onto the qubits.”
Atom researchers plan to refine and improve this capability for potential inclusion in future quantum computing architectures. It also can be useful for quantum storage or quantum RAM.
Long coherence times are essential, but many other ingredients are needed to build an intermediate scalable error-mitigated quantum computer.
Atom Computing has already demonstrated the sophistication of its qubit control system by the arbitrary manipulation of individual qubits in parallel rows and columns. The technical details of how this was accomplished are included in the record coherence time documentation.
As part of its next-generation quantum architecture, Atom Computing will be investigating other essential areas such as high fidelity single- and two-qubit gate operations, error correction, and entanglement. I assume much of this work will carry forward to its next-generation quantum processor. Since the current QPU contains 100 qubits, and there exists a high scaling potential of nuclear spin qubits, I would expect the next-generation system to have upwards of 1000 qubits or more.
In the future, it will also be helpful to see some form of holistic system benchmarking at the qubit level. Scaling is also essential, but the ability to scale is already a significant advantage of neutral atom architecture so I anticipate no problems with that.
Quantum computers have shown the potential to solve problems far beyond the scope of today’s largest supercomputers. Combined quantum properties of entanglement (a mysterious quantum force linking qubits together), superposition, and interference allow quantum computers to process multiple logical states simultaneously to generate a fast solution.
At this stage of development, there is no way to predict which technology will allow us to build fault-tolerant quantum computers with millions of qubits. The answer might be a yet-undiscovered technology, or it could be an enhanced version of qubit hardware in use today.
Despite the small percentage of naysayers, there is a very high probability that a quantum computer like that will emerge within the decade.
Note: Moor Insights & Strategy writers and editors may have contributed to this article.
Moor Insights & Strategy, like all research and tech industry analyst firms, provides or has provided paid services to technology companies. These services include research, analysis, advising, consulting, benchmarking, acquisition matchmaking, or speaking sponsorships. The company has had or currently has paid business relationships with 8×8, A10 Networks, Advanced Micro Devices, Amazon, Ambient Scientific, Anuta Networks, Applied Micro, Apstra, Arm, Aruba Networks (now HPE), AT&T, AWS, A-10 Strategies, Bitfusion, Blaize, Box, Broadcom, Calix, Cisco Systems, Clear Software, Cloudera, Clumio, Cognitive Systems, CompuCom, CyberArk, Dell, Dell EMC, Dell Technologies, Diablo Technologies, Dialogue Group, Digital Optics, Dreamium Labs, Echelon, Ericsson, Extreme Networks, Flex, Foxconn, Frame (now VMware), Fujitsu, Gen Z Consortium, Glue Networks, GlobalFoundries, Revolve (now Google), Google Cloud, Graphcore, Groq, Hiregenics, HP Inc., Hewlett Packard Enterprise, Honeywell, Huawei Technologies, IBM, IonVR, Inseego, Infosys, Infiot, Intel, Interdigital, Jabil Circuit, Konica Minolta, Lattice Semiconductor, Lenovo, Linux Foundation, Luminar, MapBox, Marvell Technology, Mavenir, Marseille Inc, Mayfair Equity, Meraki (Cisco), Mesophere, Microsoft, Mojo Networks, National Instruments, NetApp, Nightwatch, NOKIA (Alcatel-Lucent), Nortek, Novumind, NVIDIA, Nutanix, Nuvia (now Qualcomm), ON Semiconductor, ONUG, OpenStack Foundation, Oracle, Panasas, Peraso, Pexip, Pixelworks, Plume Design, Poly (formerly Plantronics), Portworx, Pure Storage, Qualcomm, Rackspace, Rambus, Rayvolt E-Bikes, Red Hat, Residio, Samsung Electronics, SAP, SAS, Scale Computing, Schneider Electric, Silver Peak (now Aruba-HPE), SONY Optical Storage, Springpath (now Cisco), Spirent, Splunk, Sprint (now T-Mobile), Stratus Technologies, Symantec, Synaptics, Syniverse, Synopsys, Tanium, TE Connectivity, TensTorrent, Tobii Technology, T-Mobile, Twitter, Unity Technologies, UiPath, Verizon Communications, Vidyo, VMware, Wave Computing, Wellsmith, Xilinx, Zayo, Zebra, Zededa, Zoho, and Zscaler. Moor Insights & Strategy founder, CEO, and Chief Analyst Patrick Moorhead is a personal investor in technology companies dMY Technology Group Inc. VI and Dreamium Labs.