T-Rex at right swipe. Genie in a bottle cap. Potter’s invisible cloak on rent. No matter how you try to explain it, nothing conjures up the sheer scale and surreal wonder of QaaS better than just saying it as it is – Quantum as a Service or QaaS.
Quantum! The word is enough to convey how elusive these machines can be. We know why quantum computers draw so much fascination and curiosity. They are way past classical computers and even supercomputers due to their ‘quantum’ advantage. Physics has manifested so brilliantly in the world of technology that it is hard to imagine that we were content working with bits and bytes just a few months back. Bits meant zero or one – in classical computing. But in the quantum world, there can be two states that exist simultaneously (0 and 1 at the same time). This happens thanks to the superposition principle of quantum physics. What follows is a tremendous spike in processing power than was possible with the 0 or 1 representation of data.
QaaS is not just another alphabet soup. Unlike servers, infrastructure, and data-center hardware that easily and brilliantly fit into the cloud model, quantum computers are too huge or complex to shrink in a cloud bubble-wrap. And yet it looks like a natural course that this giant turning point in technology had to take.
Now that qubits are in the game and now that entanglement, superimpositions, and two-state computing are not just possible, but practical; who would not want to exploit the immense speed and scale of calculations that quantum computers pack?
Except that it is not that easy to whip up a quantum computer. It’s a complex exercise and the task of maintaining the stability of quantum states is riddled with issues around hardware, super-cooling requirements, and costs.
Maybe, that’s why QaaS makes so much sense. Spare yourself the expense and headaches of maintaining a quantum computing infrastructure and just use it the way you have used infrastructure on the cloud. You have a problem or application which needs massive computer muscle? Just assign it to a QaaS facility and it will spit out quick and actionable results in the blink of an eye (almost) – without you needing to entangle yourself in all the investment and wiring.
Is it a perfect microwave for enterprises? Yes and no.
Cloud but not cloudy
Yes, there is a huge potential expected from Quantum Computing and QaaS model, which affirms Kalyanaraman K, MD, Protiviti Member Firm for India. “It is going to drive that potential to make it accessible for the benefit of the world in solving complex problems, which probably would not have been possible in the case of current classic computers.”
R Ray Wang, the Principal Analyst, Founder, and Chairman of Constellation Research Inc. unravels the positive side of this new current gaining force in the enterprise industry. “The plus side is the scarcity of expertise available in the current commercial market. Also, think of the ever-evolving improvements in technology, and the huge investment of capital.” According to him, the market is rapidly evolving and we will see QaaS as part of larger clouds. “AI and automation would be strong. And in error correction, we would witness the biggest advancement.”
Similarly, speaking on the recent announcement of the MeitY Quantum Computing Applications Lab, AWS Quantum Computing Director Simone Severini had said that there are a lot of future opportunities that need to be tapped. “It is an early stage for quantum computing and simulation applications. It sounds like a good idea to set up a lab for using quantum computing for accelerating quantum computing-led research and development (R&D) in India. This will be a service that government ministries and departments, researchers, scientists, academia, and developers can tap whenever they need it.”
Ajay Sawhney, Secretary, MeitY, had remarked that start-ups and young companies can definitely use such avenues to get started on quantum computing with a simple offering. “Quantum is at a nascent stage in the country. It is important to make access available to the young segments and give them a taste of what is possible so that they can actually experiment and build solutions around it. In quantum computing, there is still a tremendous space that we need to learn about together. Access that is simple and available for widespread utilisation is the right move for progress. We welcome AWS to the Indian ecosystem. It is making continuous investments in India with new regions and zones being added.”
According to a report by ResearchandMarkets, while classical (non-quantum) computers make the modern digital world possible, there are many tasks that cannot be solved using conventional computational methods – and that explains why QaaS can count on a growing market. The report points out reasons like limitations in processing power, or how fourth-generation computers cannot perform multiple computations at one time with one processor. Consider then what happens when a quantum computer is capable of computational feats that are orders of magnitude greater than conventional methods.
So far we have seen how parallel computing is achieved in classical computers via linking processors together. But quantum computers may handle multiple computations with a single processor. And this quantum parallelism emerges as a major difference between hyper-fast quantum computers and speed-limited classical computers.
So the market is expected to grow – or explode, as some may surmise. As per estimates by Allied Market Research, the global enterprise quantum computing market stood at USD 650 million in 2017 and can reach USD 5,853 million by 2025.
The optimisation segment is expected to exhibit significant growth in the global enterprise quantum computing market between 2018 and 2025. Interestingly, Asia-Pacific could turn out to be the fastest-growing region for the enterprise quantum computing market. This can be explained by the presence of major players and the increase in investments by government organisations.
From what Gargi Dasgupta, Director, IBM Research India, and CTO, IBM India South Asia, unravels, the hotspots are interesting. “We’ve also said that chemistry, complex optimisation problems (for industries like finance), and machine learning are main areas of focus. And we have a number of examples in these areas. But many industries are exploring how quantum computing could impact the big problems they’re trying to solve.”
As Kalyanaraman says, “All the big players on cloud computing have already started offering quantum computers for scientific research, education, building circuits, developing applications, eco-system, etc. We will be seeing a lot of enterprises working together to potentially bring this quantum computing environment to a more acceptable and compatible model with the necessary infrastructure, technologies, and skillsets to connect and commercially exploit the quantum computers through the cloud access model. As of now, the only viable model we see to harness quantum computers is through the cloud access model or making it as a service.”
Buy or rent, now or later?
Easy access and simplicity can be great pull factors for enterprises to look forward to QaaS. And yet, some details need to be ironed out before this world becomes comparable to cloud models.
Wang does not flinch from asking the business question here. “The flipside is ownership today may give some industries a competitive advantage. For example in life science, pharmacy-discovery, you could find a technique that enables faster molecule-to-market or an aerospace company can simulate material science more efficiently and they want to invest for that competitive advantage.”
Besides the alternatives that the customer side is confronting, there are questions that even QaaS vendors have not figured out yet. The pace of development in hardware areas is quite quick and slippery with quantum computing. Investing too much in one type may backfire when a new option or configuration pops tomorrow.
Hardware stabilisation does pose a challenge for QaaS providers until quantum computing matures, Wang avers: “However, not if players play it smartly.” He cites how Honeywell built its trapped iron H series. Now that should not be as much of an issue.
“Classical hardware does not pose a challenge to using IBM quantum systems over the cloud. The quantum hardware is fundamentally different from the classical computer,” Dasgupta reasons as well.
Even if we forget the ‘hardware’ parts for a while, there are other areas to be untangled. Players are also struggling with the hurdle of ‘noise’ – which is a recurrent problem with quantum computers. Approaches for reducing interference and instability are being worked out for this challenge. At the same time, they are trying to handle problems around peak-performance variations, errors, and optimisation which assume a serious connotation when a quantum computer is served on a QaaS plate.
Interestingly, the constraints themselves can be flexed as a strong route for QaaS too. Kalyanaraman says that significant constraints in quantum computing currently are its heavy capital investment, physical environment, and noise-free technology in deploying necessary technology and resources. “Bringing such large investments in the present scenario as a Capex model is far-fetched or may not be possible for several large enterprises; leave alone the mid- or small-scale enterprises. When cloud has given the organisations an OPEX model of using computing power, in no time we will be witnessing the growth of QaaS as the next level of commercial offerings by the cloud providers of the world.”
He believes that the world is yet to know more about quantum computers per se. “At this very early stage of evaluating quantum computers for commercial use, the entire technology around it is yet to be fully built in a reliable and sustainable manner, leave alone its technical challenges, economic feasibility, physical and environmental challenges, availability of skilled manpower, and related challenges it may bring upon as we progress ahead.”
A lot of players started juggling their options and constraints already. While IBM and Google are at it on the core speed and scale race of quantum computing, enterprise giants have started fleshing out the market areas in their own ways. AWS (Bracket), D-Wave Systems, Microsoft, IBM Q, Rigetti Computing’s Quantum Cloud Service, Intel, Atos, and others have their own ponies in the ring now. Atos has even devised a new metric – The Q-Score for measuring quantum performance – that, it says, applies to all programmable quantum processors. It claims that in comparison to qubits, Q-Score provides “explicit, reliable, objective, and comparable results when solving real-world optimisation problems”. It has three parameters: application-driven, ease of use, and objectiveness and reliability.
On the chip front too, there is Intel’s Horse Ridge II, a move towards scalable quantum computers. The chip, as stated by Intel, would read qubit states and control several gates simultaneously and enable faster on-chip, low latency qubit state detection. All that could boil down to the scalability of quantum computers, which would matter a lot for enterprise-level usage and for QaaS models.
One major factor to reckon here is the pricing model. Would it be on the lines of cloud economics again?
Dasgupta answers, “We have not labeled anything as ‘QaaS’, but access to all of our systems is via the cloud. And organisations in the IBM Quantum Network are IBM clients with cloud access to our premium quantum computers. Regarding pricing, we do not release any contract terms.”
Looks like the wave has started forming into a shape. Enterprises are toying with this next wave and the potential is quite appealing for companies in the finance, logistics, telecommunications, transportation, and pharmaceutical sectors. IDC has augured that 25% of Fortune 500 companies would secure a competitive advantage from quantum computing by 2023. That should be fun and fruitful. But let us not forget that even cloud models are wrestling with problems around economics, sprawl, data integrity, security, migration, integration, and skills – even after all these years.
Would QaaS be quite a party then, as simple as playing song requests with just a spin?
T-Rex at right swipe. Genie in a bottle cap.
Potter’s invisible cloak on rent. Quantum with a DJ. Sounds better?
“Chemistry, complex optimisation problems (for industries like finance), and ML are main areas of focus. We have a number of examples in these areas.” — Gargi Dasgupta, Director, IBM Research India & CTO, IBM India South Asia
“As of now, the only viable model we see to harness quantum computers is through the cloud access model or making it as a service.” — Kalyanaraman K, MD, Protiviti Member Firm for India
“There are a lot of future opportunities that need to be tapped. It is an early stage for quantum computing and simulation applications.” — Simone Severini, Director, Quantum Computing, AWS
“The plus side is the scarcity of expertise in the current commercial market. Also, think of the ever-evolving improvements in technology and the huge investment.” — R Ray Wang, Principal Analyst, Founder, & Chairman, Constellation Research
By Pratima Harigunani