Technologies of Future from the Smallest of Scales
February 11, 2021
February 11, 2021
See the full article at “Quantum Technologies: A Review of the Patent Landscape”.
Quantum technologies is an umbrella term incorporating an arsenal of technologies finding more and more applications every day. Quantum mechanics was an intellectual revolution in the twentieth century. Along with explaining previously intractable natural phenomena, it offered a new playground for engineering. We have analyzed 48,577 patents in Quantum Technologies filed from 2015 onwards captured using a comprehensive query in Relecura’s patent database (see this link for an up to date interactive taxonomized version of the patents). Contrary to common belief, Nanotechnology dominates the patent set rather than the quantum information trio (Quantum Computation, Communication, and Cryptography). Nearly half of the patents were Chinese (23,335 patents), and the US had a share close to a quarter (8,935 patents).
Overview of the Taxonomy
The most populated first-level nodes in our taxonomy are Electric Elements, Nanotechnology, Quantum Computation, Quantum Communication, and Quantum Cryptography. We did an overlap study with these nodes, and the results were not surprising at all.
Figure 1: Overlap between first-level nodes.
Electronics has been operating on the nanometer scale for decades, and both nodes (Nanotechnology and Electrical Elements) go hand in hand. The leading candidates for qubits (fundamental building blocks of quantum computers) are silicon spin qubits and superconducting qubits. Both of them leverage existing expertise and infrastructure in semiconductor fabrication (Intel has an edge in this area). The quantum information trio has a symmetric overlap between each pair. Security is baked into quantum communication, at least in theory, by the laws of nature, and advancements in quantum communication will help develop the quantum internet, an ideal way for quantum processors to connect. We will take Nanotechnology and the quantum information trio to be representative of the entire patent set from here on.
In the country-wise distribution of patents across these nodes, China keeps its overall lead. The runner-up, the US, is dwarfed by China across all nodes except for Quantum Computation, where the difference is small. From analyzing scholarly literature and news, we concluded that the US still keeps its edge in quantum computation way ahead of China, thanks to its giant corporations. China is at the forefront in quantum communication and cryptography due to its government-level agenda for domination in this area.
Figure 2: Country-wise distribution of patents across different first-level nodes.
Patent numbers are only part of the story; it takes only a few patents to bring about an industry-disrupting technology. Some sources suggest that Chinese patents are mostly of “little value” since researchers in this country patent aggressively at the expense of quality to meet their objective evaluation. Relecura quantifies patents’ quality using a custom star rating system ranging from 0 to 5 in increments of 0.5. Relecura star rating considers key patent parameters like forward and backward citations, geographical and family values, etc. We found that Chinese patents cap at 3.5 while the US has patents all the way up to 5. A normal distribution fit on the star ratings (Figure 3) arguably reveals that US patents are more valuable. The meat of a normal distribution (around 68%) is concentrated in a standard deviation length from the mean value. For US patents, it is from 1.37 to 2.81, and in the case of China, it is from 1.13 to 2.27.
Figure 3: Star rating distribution of US and Chinese patents fitted with a normal distribution.
To study this further, we plotted the forward citation network of US and Chinese patents (Figure 4). The study of forward citations is a key tool in patent landscape analysis; a patent contributing to the forward flow of information through many citations is generally considered more valuable. We also used the PageRank, the algorithm Google used to rank websites in its initial days, to compare the patents. The algorithm rewards patents connected to well-cited patents. We adapted its parameters to best perform in the context of patents where we usually don’t care beyond citations of citations. The US graph has more patent clusters with higher PageRank values, while the Chinese graph is scattered into small fragments apart from the central cluster. Though it is tempting to conclude that Chinese patents are inferior, these metrics may be unfair because of the possible variety and novelty of their inventions. We leave this open to interpretation.
Figure 4: Graph structure of the forward citations between patents. The node size is proportional to the number of citations originating from it, and warmer colors mean a higher PageRank value.
The US companies in the top 10 assignees (Figure 5) are almost centered on the Quantum Computing node with an overflow to the Nanotechnology node from their investment in the semiconductor industry. 435 of Alphabet’s 505 patents and 700 of IBM’s 876 patents go to Quantum Computing, which is not a surprise as they are the industry leaders in the field, and their business is computing. Intel has 459 patents in Nanotechnology and 280 in Quantum Computing; they are working on building quantum processors and collaborating with QuTech in the Netherlands. Northrop Grumman, one of the world’s largest defense technology providers and maker of the renowned B-2 stealth bomber, is actively exploring new disruptive quantum technologies. Their portfolio is concentrated in Quantum Computing (283) and Superconducting Devices (388 patents).TCL Corporation (the makers of Blackberry phones from 2016 to 2020), which manufacture mobiles, display panels, etc., is patenting heavily in QLEDs (704 patents) and FabricationTechnologies. BOE Technology, one of the world’s largest producers of LCD, OLED, and flexible displays, has a similar patent profile (434 patents in FabricationTechnologies and 704 patents in QLEDs). Chinese Academy of Science, the only academic institution in the top 10 assignees, has a well-balanced portfolio in quantum technologies with patents in all first-level nodes.
Figure 5: Top Assignees
LG is more swayed towards Semiconducting Devices and Nanotechnology, which is expected since its primary trade is consumer electronics. The other Korean giant in top assignees, Samsung, shows similar trends except that the numbers are scaled. Almost all of its patents come under Nanotechnology.
Though Toshiba is known for its consumer electronics products, it has diversified its business model to include IT solutions like quantum cryptography in 2020. They hold hundred plus patents in Quantum Computing, Communication, and Cryptography.
When taken alone, Nanotechnology (37,868 patents) or its child node Quantum Dots (27,526 patents) accounts for more than half of the patents we captured. Nanotechnology is the engineering of functional systems at a molecular scale. Quantum dots (QD), a central topic in nanotechnology, are nanoparticles of semiconductors with a size of the order of the Bohr radius (the scale at which particles start to show quantum behavior) of the charge carriers. QDs have size-tunable optical and electronic properties, which finds them all kinds of applications from lasers, photodetectors, solar cells, image sensors, LEDs to biological tagging, and printing inks.
QLED TVs might be the best-known example of a quantum technology in the consumer market. It is an improved backlit display technology offering a more power-efficient, long-lasting, cheap, brighter, and wider color gamut alternative to OLEDs. Samsung dominates this node with 354 patents, and they are the market leaders in QLED TVs. TCL Corporation (121 patents) and Hisense (58 patents) are in the market to compete against Samsung. Even LG (104 patents), who denounces QLEDs on their website, is active in R&D in this area.
QDs are excellent candidates for biological tagging, thanks to their size-tunable color and efficiency in capturing light. QD image sensors may save the smartphone industry from the plateauing of camera technology. For the past few years, smartphone cameras have improved significantly, but the software was doing all the heavy lifting. Despite offering industry-leading camera performance, Google’s Pixel lineup has not upgraded its sensor (SONY IMX363) for the last three generations. QD sensors provide better low light performance and higher dynamic range compared to the silicon alternatives. This technology got heated up when Apple acquired InVisage Technologies (26 patents and third in the Image Sensors node), the company that pioneered QuantumFilm technology, which allows the QDs to be sprayed onto silicon, significantly reducing the fabrication costs. An iPhone with a QD image sensor is likely to be released in the near future to claim back the smartphone cameras for hardware again.
Quantum Computing may be the most magical node in our taxonomy. IBM (700 patents), Alphabet (434 patents), Northrop Grumman (282 patents), Intel (280 patents), D-Wave Systems (231 patents), and Microsoft (218 patents) are heavily invested in this area. Quantum computing manipulates the universe’s underlying quantum nature, particularly superposition and entanglement, to speed up certain types of calculations. Quantum computers are not a competition to classical computers; so far, only a handful of algorithms that offer exponential speed up on the quantum computer have been invented. IBM presently operates quantum processors accessible through the cloud for free. Google has arguably achieved quantum supremacy (demonstrating that a quantum computer can solve a classically intractable problem in a reasonable time) using their Sycamore processor in October 2020. Most companies in this field are working towards quantum supremacy. It is likely to be achieved in the immediate future since the algorithm does not necessarily need to perform anything useful.
Quantum annealing, a significant improvement over the classical optimization method called simulated annealing, is another direction in which R&D in this field is heading. D-Wave Systems capitalizes on quantum annealing processors, rather than putting out a Turing complete (capable of computing everything computable) quantum computer. Their customer base includes Lockheed Martin, Google, NASA, etc. and their computers are finding applications in optimization, machine learning, and material science. However, we couldn’t find any conclusive studies on the performance of these processors.
For quantum computers to change the world, the algorithms have to be able to scale up. Shor’s algorithm can offer exponential speedup to the integer factorization problem, which is classically hard for large integers that the RSA cryptography system depends on it as a one-way hard problem. The algorithm’s current best record is factoring 35, and quantum computers must move from the sub-hundred qubit realm to a few thousand qubits for these algorithms to be of any help in real-world applications.
Nothing is forbidding putting together a 1000 qubit machine, but they have to get influenced by each other yet stay isolated to external influences. Quantum systems are susceptible to external noise, and quantum error correction is challenging due to the no-cloning theorem. Classical error correction algorithms mostly rely on redundancy, but this theorem forbids the replication of arbitrary quantum states. Error correction was once thought to be so hard that some scientists once believed that quantum computers would never fly. But people came up with ingenious algorithms, and we got our present state of the art, Noisy Intermediate Scale Quantum (NISQ) computers.
Figure 6: Emerging Technologies
Maybe a few thousand qubit quantum computers will be a reality someday, and they will get to break the RSA and Elliptic-Curve cryptosystems. But the caveat is that if someone records today’s internet traffic along with the public keys exchanged, he/she can decrypt the data, possibly before losing its significance. Cryptographic algorithms relying on ‘hard’ mathematical problems will get weaker over time as the development of computers catches up to relabel ‘hard mathematical problems.’ Presently, we try to leave enough headroom to accommodate these growths, but quantum computing is growing at an industry-disrupting pace. Yet, there is a quantum solution to this quantum problem.
Quantum Key Distribution (QKD) involves exchanging the public key through a quantum channel where the no-cloning theorem and collapse of the quantum state on measurement comes to our aid. In theory, it is as safe as the communicating parties exchange the keys in person that the attacker is left with dumb luck alone to figure out the key. If an eavesdropper tries to measure the message carrier’s state, the state will randomly collapse, and the receiver would know about the attack.
Figure 7: Top Assignees in Quantum Key Distribution
This technology has some political undertone to it. China is aggressively pursuing it as a political agenda, and they are at the forefront of research in QKD. All of the top 10 assignees except Toshiba and Nokia are Chinese. China has been redefining the records for terrestrial optical fiber QKD links, and they are the first to establish QKD with a low-orbit satellite (Micius) over a distance of 1,200 km. The US is not seriously considering QKD as post-quantum cryptography due to the known weaknesses in physical implementations and the cost. Symmetric key encryption systems (same key for encryption and decryption) are vulnerable only to Grover’s search algorithm, which can search through a database of N elements in a time of the order of N (square root speed up). So a symmetric key cryptosystem using a 256-bit key will be as safe as a 128-bit key system in the pre-quantum world. Plus, cryptographic systems that are unconditionally secure in all computation models (see one-time pad), if used correctly, are already available, albeit at the expense of increased computational load.
Many experts believe that the present state of the art, NISQ computers, are likely to benefit quantum physics simulations, at first. Quantum many-body simulations are hard, even for supercomputers, due to the amount of storage involved, forbidding our understanding of important problems like high-temperature superconductivity. If we could engineer a quantum system to mimic the original, the simulation problem will be turned into a measurement problem. Quantum computers are already being used like these, and the available number of qubits is well past the number of particles supercomputers can handle in many-body simulations. Plus, quantum simulations are relaxed when it comes to error corrections. Decoherence errors are often desirable as we are trying to simulate a real quantum system, not an ideal computational model.
Quantum technologies are being launched into the industry at the right time for our civilization and earth. Nanotechnology and electronics are working in conjunction to put out devices that offer better performance and energy efficiency. The onset of better lighting solutions, semiconductor devices, etc., will lighten our energy budget, giving some cushioning in the transition towards renewable sources.
If ever realized quantum computers could do more than cutting down the energy we spend on computation. We can use quantum simulators to study molecular interactions and develop efficient industrial catalysts. Through full chemistry solutions, efficient alternatives to the Haber-Bosch process, responsible for 1% global emissions, can be studied, and better battery chemistries can be invented. Quantum optimization algorithms can help us better use our resources to accommodate the growing population and climate change and someday even plan space missions.
While there is no sure or impossible thing in the tech world, quantum technology thrives on its promises, attracting investments more than ever. IBM’s Watson was introduced 64 years after the ENIAC, and we went to the moon 66 years after the first flight. Quantum technologies have already put products on the market, and as a whole, the field is iteratively improving. Government agencies and industries are globally collaborating on an unprecedented scale to materialize the promises of this field. Many of today’s billion-dollar businesses like the internet, nuclear energy, GPS, etc., were publicly funded or developed for military applications before there was a commercial market to them. If anything, quantum technology is a prodigy, optimistically set to change everyone’s life positively in the next decade at the latest.
Catch up with all of our content exploring the latest trends, topics, and thoughts from the world of innovation.