An ordinary laptop just solved a problem everyone assumed needed a quantum computer

[AlexaBliss]

Researchers used tensor networks to compress the overwhelming wave function created by hundreds of entangled qubits, and solved a problem on relatively modest classical hardware that had previously been described as impossible without a quantum computer

This is not the first time a classical algorithm has clawed back ground that quantum computing was supposedly going to own outright, but it is a good reminder that quantum advantage claims need to be checked against the best classical methods available at the time, not against naive brute force

The tensor network approach works by finding structure in the entanglement pattern that lets you represent the system far more compactly than a full exponential state vector would require, which is exactly the kind of mathematical trick that keeps eating into quantum's claimed advantage

For people tracking quantum computing progress this is a useful corrective, the goalposts for genuine quantum advantage keep moving because classical algorithms keep getting smarter about compression and approximation, it is not that quantum hardware regressed, it is that the classical bar rose

None of this means quantum computing is not making real progress elsewhere, but it does mean any specific advantage claim has a shelf life until someone finds the classical shortcut, and that shelf life has been getting shorter not longer over the past few years

An ordinary laptop solved a problem thought to require a quantum computerwww.sciencedaily.com

[Jordan89]

This keeps happening and it is honestly the healthiest thing for the field, forces quantum advantage claims to actually mean something rather than just being marketing

[Undertaker00]

Every time this happens someone posts the classical algorithm undid it headline and every time the underlying quantum hardware progress continues anyway, the two threads are somewhat independent

[Nina_20]

Right, ordinary laptop probably means a well specced workstation running for a while, not literally someone's daily driver machine

[Emma29]

Worth remembering ordinary laptop is doing some heavy lifting in that headline too, tensor network methods still need serious computational resources even if it is not a quantum device

[ArVeeDee]

Genuine question, does this mean the original problem was never actually a good candidate for demonstrating quantum advantage in the first place

[StringTheory97]

Tensor networks specifically have been chipping away at quantum simulation claims for years now, this is a continuation of a trend not a one off surprise

[Cobalt Sophie]

Probably yes in hindsight, a lot of early advantage demonstrations picked problems that happened to have exploitable classical structure nobody had found yet