The Race to Build the First USEFUL Quantum Computer

and why we don't have one yet ft. Chetan Nayak Microsoft Technical Fellow

This past May I had the chance to tour Microsoft’s secret quantum lab in Seattle and see what exactly are they getting up to in there. Did they really engineer a new state of matter? What in the world is this Majorana particle and what makes this quantum chip any different from the rest?

But first thing’s first: What the heck is the difference between a quantum computer vs a regular computer like this laptop I’m using to type this Substack article on?

Here's the first thing to know: quantum computers aren't just really powerful supercomputers. They're fundamentally different machines altogether.

Here’s what I mean:

Imagine you're trying to solve a massive maze. With a classical computer it’s like having one character who methodically explores every path - they hit a dead end, backtrack, try the next route, rinse and repeat until they find the exit. It's systematic, reliable, and painfully sequential.

Whereas a quantum computer? It's like being able to split the character into multiple versions and explore every possible path in parallel, and thereby increasing the chances of finding the right exit and quicker.

For certain types of problems, this parallel exploration gives quantum computers exponential advantages. Imagine if the maze had 300 decision points. A classical computer might need to check 2^300 possible paths (more than the number of atoms in the universe). Whereas, a quantum computer with 300 qubits could theoretically explore all those paths simultaneously.

But, quantum computers are only good at specific types of problems for example: breaking encryption, simulating molecular behavior, certain optimization problems, and pattern recognition tasks.

Whereas, classical computers excel at: sequential tasks, precise calculations, anything requiring step-by-step logic

So quantum computers aren't replacing your laptop. They're solving problems your laptop couldn't touch in a trillion years, while your laptop continues doing everything else better.

The real question isn't whether quantum computers will take over - it's when they'll become useful enough to tackle problems that actually matter. And that timeline might surprise you. Watch the video at the top for the answer to that. Chetan Nayak, one of the masterminds behind Microsoft's Majorana 1 chip, his answer about whether quantum computing is really years away instead of decades was very intriguing.

But the bigger question I had, and is the namesake of this article: Why don’t we have a quantum computer yet?

The answer is we DO have quantum computers, but we don’t have one that is actually useful…YET. But, with all of our technology, how have we not come up with one by now??

Compared to the progress of classical computers like this laptop, the quantum computers we have today are more like a proof of concept prototype - impressive demonstrations that hint at future potential but aren't practical for everyday use. They can only solve very specific problems under very controlled conditions.

So what’s taking so long?

Physics, mostly.

Qubits - the quantum version of computer bits - are incredibly fragile. Even the tiniest interference or disturbance from the environment can destroy their quantum properties in milliseconds. Current quantum computers lose their "quantumness" so fast that you can barely get them to do anything useful before the quantum effects collapse.

See, classical computers mess up about once every 100 quadrillion operations. Quantum computers? They're wrong roughly every thousandth calculation. You can't build reliable systems on that foundation.

So, why not just add more qubits to power through the errors? Just like how you would add more bits to a classical computer to make it more powerful, the same goes for a quantum computer. But, this creates a catch-22 from hell. More qubits means more power, but it also means exponentially more opportunities for things to go wrong. Plus scaling up means larger facilities to store this machinery, more than currently makes sense.

Right now the issue isn’t even the storage, it's creating more qubits. For context, IBM's latest quantum computer has around 1,000 qubits. Google has about 105. For truly useful quantum computing - like breaking encryption or simulating complex molecules - we likely need hundreds of thousands to millions of error-corrected qubits.

But Microsoft claims to be on the path to a million qubits. Yes, 1,000,000 qubits. Did they actually crack the qubit scaling issue? You can watch the full video live now at the top!

Comment below if you want the full unedited conversation with Chetan Nayak!