Fung Institute for Engineering Leadership

Op-ed: Quantum computing: the internet will soon need an upgrade

by

Op-ed: Quantum computing, the internet will soon need an upgrade

By Garey B. Fleeman, MEng ’20 (EECS)

Photo by John Schnobrich on Unsplash
The speeding up of computers has drastically changed the modern world. According to Moore’s law, the computing power of computers doubles nearly every two years (Jurvetson). Currently, transistors (the fundamental building blocks of computers) are nearing the size of atoms which means they can’t shrink any further and Moore’s law will come to a halt. The search for new computing techniques, has led to new methods such as quantum computing, which is not only on its way, but recently had its most significant breakthrough. It performs calculations in a vastly different manner than a classical computer, and this allows it to solve some problems such as factoring large numbers into prime factors. These problems are very important to many encryption schemes and solving such schemes will present many challenges. This does not mean the end of the internet, but it will certainly bring problems that will need to be solved in parallel with the development of quantum computing to keep our data secure and in the right hands. The significance of quantum computers arises from the fact that they are not simply upgraded computers; they perform calculations in a fundamentally different way. Traditional computers have been doubling in speed at rate slightly faster than every two years since the 1970’s, but they are not closer to solving large scale computational problems because they use classical bits (Roell). While traditional computers are simply made of electric circuits, where the state of the computer is completely determined by a combination of 1’s and 0’s, Roell describes quantum computers as having vastly more states by using the quantum mechanical principles of superposition and entanglement. Instead of a traditional computing bit, quantum bits (qubits) can be in arbitrary combination of on or off (1 or 0) when not being measured. That means that before the measurement, many calculations can be performed in parallel using the same qubit. In order to take advantage of single particles being in multiple states, the principle of entanglement is used to turn these states into readable measurements. Although the promise of quantum computers is great, don’t throw away your traditional computer just yet. Quantum computers are currently very difficult to make. Unlike traditional computers, adding more bits does not necessarily give more computing power. Bits must be added, and noise must be reduced. According to Jurvetson, “… The best way currently to tackle noise is to use error-correcting codes that require significant extra qubits themselves”. The computers must also be cooled to temperatures near absolute zero which is colder than the vacuum of space. Despite this difficulty, progress is certainly being made in the field. Google and researchers at the University of California Santa Barbara recently made the most significant breakthrough to date in quantum computing, described as achieving “quantum supremacy”. A calculation which would take 10,000 years on a traditional supercomputer was performed in 200 seconds on a superconducting quantum computer with 53 qubits. The task was estimating the probability that a set of quantum bit strings “resembles a speckled intensity pattern produced by light interference in laser scatter,” (Arute). While this task is certainly very narrow and specific, Arute describes it as, “heralding a much-anticipated computed paradigm.” This narrow task was a simple proof of concept, but researchers have known about quantum algorithms for tasks such as finding prime roots of numbers since the late eighties (Arute). Researchers currently believe solving prime factorization by a quantum computer to be around 25 years away (Jurvetson). To clarify, solving this problem is in fact breaking a currently working system. It may seem that finding prime roots of numbers is also a meaningless task, but prime factorization is in fact at the root of computer encryption and currently safeguards data such as your credit card numbers and other internet traffic, keeping them hidden from the wrong eyes using schemes such as 2048-bit RSA. Asynchronous encryption schemes such as RSA, which rely on factoring large numbers into prime factors are at risk. We call an encryption scheme asynchronous because these calculations are very difficult to perform, but trivial on verification. Consider the number 51. It may take you a significant amount of time to find two prime numbers that multiply to give you 51. But given the task of verifying if 3 and 17 are those prime numbers, it is very straightforward to verify that they do indeed meet the requirements (Jurvetson). Risk of encryption being broken is a huge problem but there is currently a lot of work being done and it is considered solvable. That does not mean it will be a straightforward problem to solve, however. According to Jurvetson, typical internet users do not need to worry. Methods of encryption are already being developed, which will be sufficiently difficult for even a quantum computer to crack. But, “for governments, there is more at stake”. As Jurvetson observes, “The messages they send today — between embassies or the military, for example — may well be significant in 20 years and thus worth keeping secret. If such messages are still being sent via 2048-bit RSA encryption or something similar, then these organizations should start worrying — quickly,” (Jurvetson). The joint achievement by Google and researchers at UCSB has brought quantum computing back into the spotlight. Performing a calculation in 200 seconds that would take a supercomputer — not just a regular computer — 10,000 years is certainly a significant milestone. Although the task performed was narrowly defined, it shed more light on the future of 4 quantum computing and proved that the technology is feasible in the near future. Because quantum computers work so differently from traditional computers, they can perform tasks very difficult for regular computers such as prime factorization of large numbers. The difficulty of prime factorization is the building block of many encryption schemes used to protect data sent across the internet. The quantum algorithms to perform such calculations have already been found — they are simply waiting on the hardware. New encryption schemes are already being developed, but that does not mean that the invention of a quantum computer that can perform prime factorization will not come with problems. While you probably should not start panicking about storing your credit card information in Chrome Autofill, many organizations should. Data that already exists on the internet in current encryption schemes, which will be sensitive in the future such as government communications, are at risk and will pose the largest problem.

About the author:

Garey Fleeman is a Master of Engineering candidate at UC Berkeley with a focus on Electrical Engineering and Computer Science (EECS) and a concentration in Embedded Systems and Robotics. He loves technology, entrepreneurship, and is passionate about someday using both to help change the world for the better. Connect with Garey.

References:

Op-ed: Quantum computing: the internet will soon need an upgrade was originally published in Berkeley Master of Engineering on Medium, where people are continuing the conversation by highlighting and responding to this story.