Majorana 1: Microsoft’s Work On Quantum Computing (Part 1)

On February 19, 2025, Microsoft officially introduced the Majorana 1 chip, becoming yet another competitor in the growing quantum computing business and joining other companies such as Google, IBM, Intel, and IonQ.

However, this chip isn’t quite like it’s competitors, as Microsoft claims to utilize a new type of technology, ‘topological protection’, meaning it’s less prone to problems that other quantum computing chips face. In addition, it will have high scalability in the future, and bring about never-before-seen computing power with relatively high energy efficiency in certain tasks.

So is Microsoft’s proof-of-concept device really the first step in revolutionizing quantum computing and possibly the future of tech? Or is it just some clever gimmick?

In this two part series, we’re going to understand what the Majorana 1 is all about. But before we get to that, we have to take a look at how quantum computing works by comparing it with classical computing.

What Is Quantum Computing?

Quantum computing is still a relatively new industry considering that the concept of modern classical computing was created roughly 80 years ago with the development of the first programmable electronic computers.

The emergence of quantum computing itself can be traced back to the 1980s when American physicist Richard Feynman proposed using quantum mechanics for computation. Then in 1994, Peter Shor developed a quantum algorithm for factoring (known as Shor's algorithm), highlighting the potential of quantum computation.

However, it’s hard to fully grasp the concept of quantum computing without understanding classical computing first, so we’re going to first take a crash course in classical computing.

The Concept of Classical Computing Explained

The electronic device you’re using to read this article, whether it’s a laptop, phone, or tablet, at it’s very core is made up of a simple switch with two possible positions, 0 and 1 (on and off). This system of 0s and 1s is known as binary, and is the fundamental of all computing we see today.

On the basis of binary, other more meaningful systems are created, such as ASCII (American Standard Code for Information Interchange), which uses 8-bit/8-digit binary numbers to represent English characters as well as some punctuation and special characters. For example, under the ASCII system, the word ‘hello’ can be represented as ‘01101000 01100101 01101100 01101100 01101111’, with each string of 8 digits corresponding to each of the six letters. However, when your computer is actually reading binary, there aren’t spaces (or placeholders) in the code, so it looks more like ‘0110100001100101011011000110110001101111’. As the computer knows that eight digits correspond to a single character, it automatically understands where the binary corresponding to the last character ends and where the next binary starts, eliminating the need for placeholders and making it has simplistic as possible.

But the problem with systems like ASCII is there are only so many characters that it can differentiate. Doing the math, there are only 256 (two to the power of eight) different combinations of 0s and 1s, meaning that it’s far from enough to show all global characters such as Chinese, Hindi, French, etc. So that’s why there are more sophisticated and complex systems like Unicode have been created, aiming to include all global characters from different languages, symbols, and emojis. They use longer binary to represent a single character, with various encoding forms such as UTF-8 (1-4 bytes per character), UTF-16 (2-4 bytes per character), UTF-32 (fixed length: 4 bytes per character). This solves the problem of having limited ability to represent all characters in systems such as ASCII.

The Hardware of Classical Computing Explained

The binary system is the software of classical computing. However, that’s only half of the picture, as there is also the hardware that supports the classical computing. As the demands of computation have increased throughout the decades, there have been more specialized pieces of hardware emerge to deal with these demanded tasks.

At the heart of a computer, there’s the CPU (Central Processing Unit), which is great at general things like running the operating system, everyday computing, etc. Since it’s somewhat official release with the Intel 4004 in 1971, it not only marked the beginning of the microprocessor era, it also became essential at the tasks it’s known for today.

In 1985, the first commercially successful FPGA (Field-Programmable Gate Array), the XC2064, is released by Xilinx. These pieces of re-configurable hardware could be programmed to perform specific tasks.

However, in the late 1980s and early 1990s, there was growing demand for advanced graphics in applications such as gaming, CAD (computer-aided design), etc. So NVIDIA released the GeForce 256, and that’s how the term ‘GPU’, which stands for Graphics Processing Unit, was popularized. It’s importance has been again concreted with the growing popularity of artificial intelligence, as GPUs are exceptionally good at parallel processing, scientific computing, and deep learning, thanks to its origin in graphics processing.

2011 saw the release of AMD’s APU line, which combined a CPU and a GPU for high performance while maintaining power efficiency. These Accelerated Processing Units have since skyrocketed in popularity on portable laptops.

In 2016, Google introduced an accelerator called the TPU (Tensor Processing Unit). This popularized the usage of the NPU (Neural Processing Unit), which specialized processors for neural network computations, with other companies adding custom NPUs into their products. For example, Apple added the ‘Neural Engine’, while NVIDIA released its Tensor Cores.

5 years ago in 2020, the term DPU is introduced by companies like Intel and NVIDIA to refer to processors that are optimized for data processing, mainly in data centers.

Now you may be asking, then what are computer processors, like Apple’s M series chips and Intel’s i9? Well, these chips or processors can be seen as a single component that encompasses all necessary specific pieces of hardware. As an example, the Apple M1 chip includes both a CPU, GPU, and a Neural Engine, as well as unified memory that these processors share.

The Concept of Quantum Computing Explained

Now that we have a basic understanding of classical computing and the hardware that supports it, it’s time to turn our attention to quantum computing, which is both similar in some ways but also has a few key differences at the same time.

A quantum computer also utilizes 0s and 1s just like any other classical computer. However, they’re different as they utilize some magic, particularly spooky action at a distance and Schrodinger's cat. The first being Albert Einstein’s term for describing the concept quantum entanglement, and the latter a thought experiment designed by Erwin Schrodinger to understand quantum superposition. Both entanglement and superposition are terms that belong in the realm of quantum mechanics (or quantum physics). Quantum mechanics is a theory in physics that describes the behaviour of matter and energy at the smallest scales, such as at the atomic scale or even subatomic scales (electrons and photons). As one of the fundamental pillars of modern physics alongside Einstein’s theory of relativity, quantum mechanics is a very important concept that is still being explored by physicists yet we know not much about.

But what we do know can already be used to fundamentally change the world, particularly by applying quantum mechanics to computing (a.k.a quantum computing).

The term superposition is particularly crucial to quantum computing. Not only can the switch be in the 0 or 1 positions, it can also be any other position, for example, in between. However, when you observe where the switch is positioned, it collapses (chooses) to be either a 0 or a 1. This unique property of the bits of quantum computers is why they have a special name, the qubit. A formation of qubits can be integrated into a new type of processor, the QPU, or Quantum Processing Unit.

In addition, qubits can also be entangled with other qubits, a process known as quantum entanglement (it can also be done on other particles as well). Basically, if you observed the value of one qubit, the other one will also collapse its superposition and show a correlating value. The ‘spooky action’ in all of this is that no matter how far apart, and how isolated the two are, they still manage to ‘communicate’ and show a correlating answer.

For example, Ball A and Ball B can be either black or white, each is enclosed in a container where their color can’t be observed, and they’re entangled. If A is white, then B will also be white, and vise versa. Now, take Ball B to the far end of the observable universe, while Ball A remains on earth. In this moment, if we open the container of Ball A and observe it’s white, then Ball B will instantly become white as well, regardless of the distance or matter between them. From the perspective of common sense, the only way that this phenomenon can happen is that when you observed Ball A, the information of its color was immediately transmitted to Ball B in a period infinitely small. To achieve this, the speed at which the information was transmitted would have been faster than the ‘speed limit of the universe’ (the speed of light), however, it’s already a well-established fact that nothing (including useful information) can be transmitted faster than light. That speed limit is there to ensure the principle of causality (cause happens before the effect) is respected, otherwise the effect may be seen before the cause has.

So does quantum entanglement violate the principle of causality? On the surface, it seems so. But that’s not actually the case. You see, you can’t decide before hand whether Ball A would be black or white (it’s random), so you can’t use quantum entanglement to tell the observer of Ball B anything useful. In this case, information is transmitted faster than light, but there wasn’t any useful information transmitted, thereby respecting the causality principle.

Using superposition and entanglement, quantum computers can complete certain tasks at incredible speed, including optimizing problems, quantum simulations, research purposes, machine learning, as well as cryptography and cybersecurity.

Quantum computers are basically like even more specialized and powerful processors that theoretically will do much better than classical computers at the previously mentioned tasks. This means that there is a lot of potential in store for quantum computers. From the development of drugs and medicine to the training of highly-sophisticated artificial intelligence models, quantum computers are poised to do a lot more.

Conclusion

But that’s only the concept of quantum computing. Like classical computing, quantum computing also has hardware that supports its processes, and before Microsoft announced the Majorana 1, there were mainly two types.

However, that will have to wait until next week. Thank you for reading this article, don’t forget to smash that subscribe and share button, and remember to come back next week to see part 2.