Folks, the quantum trade is already on Wall Street’s screen, but the boys can’t seem to agree on when this potential tool of doom actually becomes useful.
Though to be fair, Goldman Sachs (GS) once looked early in the race. I mean, just three years ago, the bank hired a small group of scientists and worked with Amazon (AMZN) to test whether quantum computing could help wealthy clients get stronger portfolio returns.
The test was kind of a smack on Goldman’s face, as they had to find out that the algorithm would need millions of years to finish the task. The computer would also need at least 8 million logical qubits, which are protected quantum bits used to build a reliable machine. Today’s systems still have fewer than 100.
Banks are chasing quantum gains as the hardware still falls far short
Goldman later cut most of that team during a wider cost-cutting round. JPMorgan Chase (JPM), meanwhile, went the other way, keeping more than 50 physicists, computer scientists, and mathematicians working on optimization, machine learning, and cryptography.
Some on the Street think quantum will be the next big computing trade after artificial intelligence, while others are not ready to spend heavily on a tool that still has limited use in real business.
Tech and market experts say quantum computing could help with drug research, machine learning, finance risk models, and other hard problems that normal computers struggle to solve.
The issue is the clock we’re working with. Useful quantum systems are still seen as years away since they use physics such as superposition and entanglement. A normal computer works with bits, which are either 0 or 1. A qubit, short for “quantum bit,” can exist as a mix of two states before it is measured. When the machine handles qubits in the right way, wave effects can raise the chance of getting the needed answer.
A large quantum computer could run some calculations far faster than a classical computer; it could also help physicists run physical simulations and break some common encryption systems. Another super interesting angle to the story is Xanadu Quantum Technologies, whose founder, Christian Weedbrook, became a billionaire within literally 6 days of the company going public.
Christian’s stake in Xanadu was valued at about $1.5 billion by midday Friday after the company’s value more than tripled during the week, and Xanadu closed at $31.41 on Friday, up by 251% on the weekly charts, per data from Google Finance.
Xanadu says it plans to build one of the first quantum data centers by 2030, and it uses photons, or light particles, sent through fiber-optic links.
Then, we have the most valuable company on earth (Nvidia), which released open-source artificial intelligence models on Tuesday to support research in quantum computing.
Google lowers the bitcoin threat estimate as exposed wallets face the bigger risk
Now let’s talk about the elephant in the room: Bitcoin. But first, a trip down memory lane, all the way to 1994, when mathematician Peter Shor created Shor’s algorithm, a method that can break the trapdoor behind some cryptographic systems.
Peter’s algorithm solves the discrete logarithm problem efficiently. A classical computer would need longer than the universe has existed for some versions of that math. Shor’s method handles it in polynomial time, where the difficulty grows slowly as numbers get larger.
The algorithm has been known for more than 30 years. Bitcoin still works because no one has built a quantum computer with enough stable qubits to keep coherence through the full attack, but we wonder:- how many qubits would be enough?
Previous estimates had pointed to millions of physical qubits, but last month, Google (GOOGL, GOOG) released an investigative report that reduced that number to fewer than 500,000.
The paper also laid out a more direct attack path. Part of Shor’s algorithm depends only on fixed elliptic-curve data. That data is public and the same for every Bitcoin wallet. A future quantum machine could do that part early and wait in a ready state.
Once a public key appears, either in the mempool during a transaction or on-chain from an earlier spend, the machine would only need to complete the second stage.
Google’s report estimated this part will take about nine minutes to be done, whereas Bitcoin’s average block time is 10 minutes, so that gives a potential attacker a short window (41% to be precise) to calculate the private key and submit a competing transaction that sends the coins somewhere else.
The larger issue is already sitting on the blockchain, where 6.9 million bitcoin, roughly one-third of the total supply, is held in wallets where the public key has already been exposed forever. Those coins face an at-rest attack. But again, who knows when the danger will actually get here?
