We posted a short News article about Honeywell Quantum Solutions merger with Cambridge Quantum Computing earlier this week, but we wanted to find out more about the motivation, benefits, and opportunities of this newly announced plan to merge the two organizations, so we talked with Tony Uttley, president of Honeywell Quantum Solutions and gained a few insights into the merger.
The first thing he pointed out is that Honeywell Quantum Solutions (HQS) and Cambridge Quantum Computing (CQC) have been working together for over three years. In fact, CQC was the very first outside customer to have access to the HQS ion trap machines and even today they are the largest external user of monthly computer time on the current generation of H1 machine. The two teams combined will have one of the largest quantum teams in the industry with 320 employees today and an estimated 350 employees when the transaction closes later this year. And a large portion, roughly 200 will be scientists and engineers with about 120-130 holding PhD degrees.
Although the two organizations have worked together for some time as vendor-customer, Mr. Uttley made some interesting observations of how there will be increased synergy when the two are together. One emerging concept we are starting to see in the quantum industry is that of co-design. If a hardware and a software group work very closely together, it may be possible to develop an application specific solution that is optimized to solve specific problems. Perhaps a custom application specific chip could be designed specifically optimized to run algorithms for certain things, such as computational chemistry or materials science.
Another area where it may require a hardware and software team to work very closely together is with very low level pulse control of the qubits. Although all of the hardware vendors have their own software that will calibrate the qubits and figure out the correct sequence microwave pulses or laser light to implement their native gates, there could be opportunities to have the machine perform even better if the software group developing the higher level algorithms has access to these low level controls and optimizes the qubit operations for their algorithms, perhaps even creating new types of native gates. IBM does have a capability called OpenPulse and Rigetti has their own version called Quil-T. But utilizing this type of pulse level control is very complicated and requires a lot of support and transparency from the hardware vendor so they can provide the user with a detailed understanding about the hardware operation. So having engineers from HQS and CQC on the same team can make this easier and allow the two together to make better use of this capability.
There are other non-technical aspects of this merger that we think may benefit the new organization. The first is money. It’s our belief that in the coming months, quantum providers will need to move into the next phase and start scaling up rapidly. Where only 1 or 2 systems may be adequate to provide potential customers with platforms to do initial beta testing, as customer start getting serious a vendor may need to build more machines and make them available both for job capacity as well as user availability purposes. As the number of machines, the size of the machines, and the number of customers expand a vendor will need to spend money to expand their capabilities in order to keep up.
As part of the arrangement Honeywell Corporation will provide the new company between $270 – $300 million of cash. But the question arise whether that is enough? For comparison, in the IonQ SPAC deal, IonQ will start off with about $600 million in cash to fund R&D and commercialization activities. The advantage of having an independent company as proposed in this HQS/CQS merger is that it does open the door for a future IPO which would allow them to obtain additional public funding. Although Mr. Uttley couldn’t provide any predictions on if or when an IPO could happen, he did indicate that “every option is on the table”. One of the more interesting things we heard is that in previous months outside investors would contact Honeywell Corporate and express interest in making an investment specifically into the HQS division. Of course, that type of arrangement is not possible when the division is inside a $30 billion corporation, but it certainly is possible when that division is spun off to be its own independent company. Our belief is that there is significant general public interest in making investments in quantum companies, so we believe an IPO could be quite successful if the new company chooses to do it. But we should point out that after the merger Honeywell Corporation will still own about 55% of the new company, so Honeywell will be able to share significantly in the new company’s success.
One observation of our own is that this new structure will also provide advantages from a sales perspective. Although Honeywell Corporation is a great company, its existing business areas of aerospace, building technologies, performance materials & technologies, and safety & productivity solutions are probably not as synergistic as it could be to sell quantum solutions to an IT manager or data scientist. In contrast, companies like IBM, Microsoft and AWS sell their existing classical computing products and services to IT managers, data scientists, logistics managers, etc. on a daily basis and they have been doing this for many decades. Having a standalone company will afford them the opportunity to develop their own sales channels. In addition, we do expect there will be incremental business as CQC introduces customers to HQS and vice-versa.
Certainly one possible issue is whether there could be reluctance from other third party software companies to work with the HQS side or other hardware companies to work with the CQC side. We asked Mr. Uttley about this and he stressed that they want to have extensive collaboration with the rest of the industry. He pointed out that CQC t}ket> software already supports a very wide range of different hardware platforms and they absolutely want to continue doing it. They are striving to provide the best possible solution for their customers and it’s likely that no single platform or technology may be the best at solving every problem. A customer may want to explore to see which technology may be the best for their particular use case and HQS/CQC will encourage that. What’s more, it is possible that the best solution may consist of a hybrid/solution that may team a superconducting based machine with an ion trap machine to execute different portions of the algorithm and they would certain support exploring that approach.
There won’t be much change in the company’s locations or facilities. The U.S. headquarters will be in Broomfield, Colorado and the UK headquarters will be in Cambridge. There will be significant expansion as they staff up and they are installing a second H1 machine at a Honeywell facility in Golden Valley, Minnesota. Finally, we asked about the status of the H2 computer which will be the successor to their current machine. It is based upon a racetrack type topology and you can get an idea of its topology from a blog that HQS posted last year. Although Mr. Uttley didn’t provide specifics or pre-announce the machine, he did indicate that they already have a version of it working on their lab and it is under test. So we will need to wait to find out more about the H2, but we don’t expect we will need to wait for very long.
June 9, 2021