Quantum Computing is a form of computation that leverages the abilities of bits when they are in a quantum state to perform calculations. Classical computing, what we’ve all learned and become accustomed to, has bits that equal either 0 or 1 at any given time. From there, you can combine 8 of them to turn them into a byte, store data through the 0/1, or use AND/OR functions to perform some type of calculations. Quantum Bits (qubits), on the other hand have many more types of abilities (superposition, interference, and entanglement). Superposition allows a qubit to become both 1 and 0 at the same time, and entanglement allows for 2 qubits to be linked so that they match their state at any point in time in any part of the world – Einstein referred to this as “Spooky action at a distance.” Understanding the technology of quantum is extremely challenging, which is frustrating because as humans we’re weary of new technologies without being able to understand them – “is this a fad or is this here to stay”. How many of us heard of Blockchain and failed to understand it during the first read through of articles? Once you grasped the concept, you understood the use cases, and began believing in the technology. Quantum is extremely similar. You’ll probably read this blog and struggle to understand the technology; I don’t blame you as I’ve read 30+ articles in preparation for this blog and still don’t trust myself to explain the technology to others. However, I hope that over the course of reading this blog, you’ll begin to understand, the implications of quantum are widespread and far reaching, and it is worth a second look at before moving on.
Quantum computing is on the verge of blowing up. There’s a race between IBM, Google, Honeywell, IonQ, and many others to build the latest and greatest computers using the greatest number of cubits. IonQ announced a computer with 32 cubits in October 2020, IBM released the hummingbird chips that can offer 65 cubits in August 2020, and just 2 days ago, IBM announced the eagle chip which can offer 127 qubits. Their road map includes a 433-cubit chip in 2022, and a 1121-cubit chip for mass production in 2023. While long ways away from materializing, this is feeling very similar to the initial tech craze of the 50’s where Moore’s law began to shine. Recently, we keep hearing that Moore’s law is dead, but has anyone bothered to look at the speed of advancements physicists have made with quantum computing?
Two graphs above show the theoretical improvements: on the left is an exponential function, and on the right is a linearized version comparing classical vs quantum computing. Notice the “Particles Known Universe” line on the right graph. That’s the theoretical limit of classical computing. When people call out that Moore’s law is dead, this is what they’re referring to. We’re becoming closer and closer to that theoretical limit based off currently capabilities. Quantum is changing all of that and throwing Moore’s law and other concepts on their head. For example, a Chinese quantum computer was able to perform certain tasks in 1.2 hours that would take the current best super computer at least 8 years to perform.
Best use cases for Quantum Computing include optimization, AI, chemistry, integer factorization, searching and indexing, and scenario simulations. I’ll discuss how some of these are used and their implications for cryptography.
Integer Factorization: Roughly half of all cryptographic encryption is doing through some type of mathematical function where you multiply large 300+ digit prime numbers together. That initial math calculation is fairly quick and could be done in split seconds but finding the 2 factors of a large 600+ digit number is extremely time consuming and almost impossible in current situations. With enough qubits and a low enough error rate, quantum computing is expected to nullify and break most public shared key encryption technologies currently available and widely used in the world. This has been anticipated by the cybersecurity community, and there’s an entire field dedicated to post quantum cryptography.
Search and Indexing: Another proven use case is the ability to search within an unstructured area of information. Using Grover’s algorithm, a publicly known algorithm that’s able to perform unstructured searches, physicists can perform searches in seconds instead of years. A common use case is a password cracker for symmetric passwords (note that above we discussed asymmetric). Through the ability of performing faster calculations of integer factorization AND performing unstructured searches, the majority of currently used cryptographic calculations are deemed broken and unusable overnight. This has massive ramifications for the way that we use the current form of the internet and the way our data and privacy is maintained.
While looking at drawbacks, we can see three that are most prevalent. The number of qubits, the prevalence of quantum algorithms, and the error rates of quantum computations. Up until now, physicists have focused on increasing the number of qubits which is steadily increasing. Many experts theorize that the total number of qubits will need to top 1000 before classical super computers will be consistently outperformed which is anticipated to come in 2023 by IBM’s roadmap. Quantum algorithms are less of a concern as they have already proven fruitful as they can be built out on classical systems before qubit systems were even available. The real hurdle of this technology is reducing the error rates to the same level of classical transistors. IonQ, as I mentioned above, is selling computing power with 32 qubits, which pales in comparison to IBM’s offerings. However, the distinction is that IonQ has focused on lowering the error rate of their products to produce a more consistently efficient technological capability compared to IBM who is focusing primarily on increasing the total number of qubits.
Quantum is more feasible now than ever before. We’re seeing more and more companies throw more money at research and discovery of quantum computing technologies. While most are still proof of concepts, there are others, like IonQ that sell services and accessibility to their quantum computer via cloud services AWS and Azure. This is definitely a long play (5-10+ years) before it becomes feasible for server or company applications. However, the earlier we’re exposed to it, the more likely we can ride the wave instead of sitting by and awaiting for technology to move past us.
Update on 11/24 @ 8:30am:
As I continued to read more about quantum computing in the days after I posted this, I realized I made 2 omissions which have big impacts. 1 is talking a bit further on the implications of what this means outside of encryption, namely modeling larger more complex molecules for pharmaceutical companies which has big implications for their R&D and what types of drugs come out in the future and the speed of advancements. If Quantum is able to scale, pharmaceutical companies will be able to lease quantum computing via the cloud and run massive simulations on molecules and effectively discover what effects drugs may have on potential molecules at much faster speeds, which would lead to faster discovery times for drugs and less R&D.
The second big thing is taking the encryption piece a step further by looking at the potential effects that quantum will have on cryptocurrencies and blockchains (last 2 articles linked below discuss this). All blockchains are based on asymmetric cryptography, which is much weaker than symmetric cryptography. As discussed above, once/if quantum is able to scale at the rate anticipated, it will break essentially all asymmetric cryptographic systems and the majority of symmetric systems.
As a result, all current blockchains are vulnerable to this technology. This is why there is a big push for ‘post-quantum cryptography’ to figure out cryptographic protocols that can withstand the rise of quantum computing. This will likely be a fruitful exercise and may give rise to the next RSA (company that founded the current standard for asymmetric encryption). However, the real challenge that blockchains and cryptocurrencies face is that they are decentralized and most have no authority or governing body who can make the decisions to change the encryption standards to the post-quantum cryptographic standards that will come about. This will be a litmus test for all cryptocurrencies; if there is no governing body, how will there be a consensus to switch over to the new standards before their cryptographic standards and blockchain is easily broken?