Avoiding the quantum winter
By Nick Farina
Since Andrew Schoen's excellent 2016 post about investing in quantum computing, I've noticed the field of quantum computing has changed, but the dialogue about professional investing in the field has stayed relatively static.
Now that we have moved into the NISQ era, the field faces the possibility of solving real problems (and generating real revenue) within the next 3 years.
The word “possibility” denotes both opportunity and danger. For those relatively unencumbered by the pressures of corporate cycles, quantum computing (and QIS broadly) offers exciting prospects for both basic science and advancing human progress.
However, interpreting “possibility” as “near-term revenue” is dangerous. The QC community calls this the “quantum winter,” as described here by Scott Aronson.
The tragedy of a potential quantum winter is that it would be self-inflicted. At the moment, there is a healthy mix of funding and skepticism in the field. This allows scientists to make progress at a good clip, without undue pressure.
But, historically, when the possibility of wealth creation has been on the table, we’ve seen that humans can’t have nice things. The gold rush ensues, and things get messy.
It doesn’t have to be this way. We can commercially explore the field of QC by investing capital now for a return within 5-10 years.
But to do so, we need to be creative, and as investors we should listen to the scientists and engineers at the heart of the field. What do they think?
"No matter how hard people try, it is not possible to put a timeline on how QC will develop on the hardware or software side. Therefore, if your research is working on either building a chip or writing innovative algorithms, the right investors need to be excited by the huge potential but not have strong expectations of near term revenue, nor can they expect teams to hit every milestone on a precise timeline. In all cases, it's imperative that the academic (in either hardware of software) be entirely honest with their investor about timeline, and never cave into pressure from non-scientific investors who try and "shoehorn" them into finding revenue in the near future. Those are not the right investors for QC, and scientific founders should be wary of them.”
Johannes Pollanen, PI of the LHQS and Co-Founder of EeroQ Quantum Hardware
With this mindset, I’ll quickly explore a few keys to reconciling science with commercial investment, with the goal of a permanent quantum summer.
A funding model for a very uncertain timeline
Ultimately, I agree with those who think the field will reach a commercial tipping point and explode within the next 8 years. But the success of QC can’t simply be “forced” to fit an investor/fund cycle driven timeline, nor can it be predicted with any type of precision [1]. To illustrate this, I teamed up with recovering physicist and aspiring QC investor Stephen Hemmerle to create this sophisticated chart :
Why the uncertain timeline?
A quantum computer must be operated in a sweet spot of physics. These systems must be interactive enough to perform the operations and computations, but not so interactive that it breaks the quantum nature of the qubit.
Because of its quantum nature, there will always be some interactions that cause errors in the system. Luckily, there are possible ways to correct these errors, but they are experimentally difficult and limit your computation. As materials, techniques, and algorithms develop, we can push for better computing ability. However, regardless of the possible implementation method for quantum computing, it remains a hard problem.
Does this mean we face a ten year (or longer) wait for useful computation then? Perhaps not. Physicists are clever, and there are a lot of future “tricks” to be discovered, both in the hardware and software.
It could be 10 years, but then again, it could be possible that someone successfully creates a new material that drastically improves their qubit’s ability. It could also be true that someone develops a groundbreaking algorithm that generates value on a true NISQ device. There is simply no way to predict how or when these breakthroughs will come, which makes it difficult when exactly the revenue in quantum computing will start.
It’s for this reason - not pessimism or optimism, but rather simply lack of certainty - that I believe quantum computing funding should be entirely devoid of revenue-generation pressure until we are at a different place in the cycle of development.
Without this kind of capital, the field is essentially banking on a Hail Mary breakthrough to avoid the quantum winter. We could get lucky, but it’s more likely such a mentality will lead to multiple QC startups crashing and burning.
It will happen because the near-term traction promised in QC decks won’t materialize. When this happens, investors will become frustrated, as they have been for many earlier hype cycles. Think about Jibo, or Theranos. These are, unfortunately, common stories.
Does patient venture capital exist?
Yes.
I’ll focus briefly on biotech as a comparable model, though there are other examples of how patience and venture capital can merge, such as robotics, classical hardware, autonomous vehicles, and more.
While “everyday” startups are indeed expected to show customer traction on a seed round today, biotech is an example of one field where these startup rules don’t apply. Biotech is a hot field in today’s investment world and involves an arguably more capital-intensive and early-revenue-light financial model.
Biotech requires tens of millions before revenue and is sometimes priced using a model called rNPV that accounts for cost, risk, and time. Time is the key. Absence of an immediate return does not render the deal un-investable. Such a model has many parallels to quantum computing.
Far from shying away, investors are pouring capital into biotech. The sector was second only to software in deal volume and dollars invested in 2018 [2].
We’re also seeing some new venture funds announce 15 year terms, rather than the standard 10 year terms.
For those new to venture capital, this means that the managers of the fund have 15 years to create a return to the investors in their fund (who are called Limited Partners, or LPs).
This means a VC with a 15 year term is under less pressure to make money, and—you guessed it—their portfolio companies are under less pressure to rush their work. Pressure is the enemy of many things, including the creativity needed to create value in the NISQ era.
As David Ferguson (former Head of the Pricing and Strategy Group for PwC) puts it: “I made initial investments in QC without expecting a near-term return. This is hard stuff, and the risk is high. It doesn't fit the typical VC investment models very well, if at all, because it's nearly impossible to tie to a 10 year fund cycle, even if you're at the start of that fund.”
This does not mean QC investors are interested in smaller outcomes. They simply have more time for them to come together.
In addition to longer funds (or funds at the very beginning of a 10 year cycle), other investors particularly well-suited to QC are corporate venture funds, high net worth individuals, and family offices, all of which generally put less pressure on timeline, because their capital is ‘evergreen’ and not tied to a specific fund cycle.
How much capital is needed for QC?
The surprising fact is that QC investors don’t need billions under management. This is due almost entirely to the fact that QC often does best with small teams and does not need expensive infrastructure (offices, sales, etc.) for the near term.
For QC software startups, this means the expense profile will be similar to other software startups, with less revenue but potentially fewer expenses.
However, for quantum hardware companies, there is no getting around the need for very expensive equipment such as cryostats, high-end microwave equipment, laser and vacuum systems, etc.
If done in strategic partnership with a university, such as we did at EeroQ Quantum Hardware (or as done by others like IonQ), the expenses can be well-controlled, as long as the startup is comfortable with the rules of university sponsored research and tech transfer, including overhead and potential sharing of IP.
On the other hand, if operating as a standalone company, such as Rigetti or Atom Computing, considerable investment (between $5-20MM) in equipment is required, but the equipment itself is a capital asset.
Still, the R&D of QC hardware in the NISQ era is not billions, or even hundreds of millions.
The value investors can bring to QC
OK, so we’ve covered the time and capital required. Now, let’s come back to the question of what private investors have to offer the academics creating value in QC right now.
Because, at a baseline, academics can access capital without revenue timeline pressure from grant funding, and already have many of the skills needed to be successful founders.
For more insight, I inquired again with Professor Pollanen:
"Federal grants remain the lifeblood of scientific research, but they are becoming harder to come by and can require hundreds of hours of work for each proposal application. For NSF grants, the hit rate is about 1 in 20. Seeking patient, private capital in appropriate amounts can be a really efficient way to boost certain promising research directions. In my limited experience, you can waste as much time pitching VCs as applying for grants. However, if done right, it can take much less time and offer a path with large amounts of capital (as you hit technical milestones) and added flexibility that is complementary to grant funding.”
Okay, but what else can VCs offer asides from, well, faster money?
“It's important for commercially-aimed QC research to have focus, in addition to patience. Grants can be patient, but sometimes do not hold the PI as accountable to specific milestones as is ideal for working towards commercial QC. Private investors can provide a good balance here, helping the research stay laser focused on the project goal, without necessarily adding unrealistic pressures. Plus, having an industry collaboration helps us to bring talent into our research - and gives them incentive to add value - which can be harder in a purely academic setting. There are so many brilliant scientists outside of universities, and it's helpful to be able to bring in this kind of talent when it fits."
Here, the full picture emerges.
There is indeed a real need for private investment capital in the field of QC at this time, because it not only provides a more efficient way to acquire funding (when academics need every extra minute they can get), but also provides other means of concrete value to scientists seeking mid-long term commercial value.
We can summarize the above with a few key rules for avoiding a quantum winter:
Encourage your portfolio companies to spend all of their capital on R&D until the field is ready for sales, even if this is in 3-4 years.
There is no faster way to kill a startup than by over hiring on the sales side, renting fancy office space, and building a complete corporate infrastructure… before your product is ready to be sold. By keeping all non-R&D costs to a minimum, companies will be able to focus on creating sizable mid-term value, without the expense and pressure of attempting to build a “real business” at the present moment.
Have at least a 10 year horizon for return on capital.
It's possible, and maybe even likely, that QC startups will begin to be sold at absurd multiples within 3-5 years on hype alone, but betting on that is not a sound investment strategy and may encourage the kind of “quick flip” exits that are unhealthy for startups and the field broadly.
Do not compromise on team quality.
At this point, capturing mid-term and long-term value is all about talent. There is likely little to be gained by investing in second-tier teams. Be patient, cultivate a trusted network of QC experts, and only invest when you have confidence in the talent of the team. Importantly, I don’t think this means investors need to understand the technology perfectly. For many, that may simply not be possible, given that the field is on the cutting edge of innovation. Venture investing has always been, at a fundamental level, about gut calls on founders who radiate brilliance and persistence. These folks are plentiful in QC today, but there is a finite amount of them.
Care about the possible impact quantum computing can have on the world, hopefully for the positive.
David Ferguson notes he’s “excited about the potential to crack some of today’s unsolvable problems, including factors in climate change, disease treatment, and problems that don't fit within classical computing solutions.” If you are driven by this aspect of QC, the wait won't seem so long.
Be flexible in your expectations about how the companies you back will unfold and are willing to let scientists and engineers take the driver's seat.
I am an investor and founder. I am not a physicist. I wish I could say that I and other investors have lots of value for the scientific founders who will be creating value in quantum computing, similar to how a great SaaS or consumer VC would have to offer. We don’t. We do have some value, covered earlier in the post, but not enough to be the driver of timeline.
So, by all means, invest in quantum computing, or take investment from a VC or angel if you’re a founder, but within the above framework, in order to avoid the worst of a quantum winter. It might be a common phenomenon in new fields, but there is no reason it needs to happen to us, too.
Thanks to:
My editor Laura Cherry, and Andy Triedman for reading an early draft and helping me evolve my thoughts on this topic.
Author Bio:
Nick Farina is the CEO and first investor in EeroQ Quantum Hardware, which he founded in 2016 with Professor Johannes Pollanen of the Laboratory for Hybrid Quantum Systems, and Faye Wattleton (former CEO of PPFA and Board Director of Columbia University). At EeroQ, he represents the company on the United States Quantum Industry Coalition, and as a sponsor of unitary.fund. He serves as an advisor to multiple Fortune 500 companies, VC firms, and top-tier consultancies on the development of quantum computing for business.
Nick is an angel investor, founder of EOC (leading provider of software and outsourced management solutions for higher education), former Board Director of Chicago Access Corporation (Chicago Public Television), and was an Advisor & Investor to Bering Science, a data analytics platform for public and private scientific research. He has extensive experience, both successful and unsuccessful, in founding and financing early stage tech ventures. He holds a BA from Haverford College, with a year of research in Meiji-era Japanese history at the University of Oxford (Mansfield College).
References:
[1] I should note there are some exceptions to the “no near-term revenue in QC” rule, particularly companies like Bluefors, Quantum Machines, Q-CTRL, and others that are selling ready or nearly-ready solutions to QC builders. In these cases, near-term revenue may be possible, and for some companies, it is perhaps already a reality. My focus for this piece are QC companies that aspire to sell hardware or software solutions to corporate end-users.
[2] https://pitchbook.com/news/articles/18-charts-to-illustrate-us-vc-in-2018