Google has unveiled a new chip which it claims takes five minutes to solve a problem that would currently take the world’s fastest super computers ten septillion – or 10,000,000,000,000,000,000,000,000 years – to complete.
Its director, Michael Cuthbert, told the BBC he was wary of language that fuelled the “hype cycle” and thought Willow was more a “milestone rather than a breakthrough”.
Frustrating that the article seems to pretty heavily focus on pop sci talk about quantum rather than actual point of what they’ve made. The actual news being methods that “present device performance that, if scaled, could realize the operational requirements of large scale fault-tolerant quantum algorithms.”
Of course its fucking Google.
RIP most encryption?
No, you’d need at the very least 2000 logical qubits to break some relatively outdated encryption, this one only has 105 physical qubits (and at their current rate they’d need over 1000 physical qubits for every logical qubit)
and even if you had that, you might still run into other problems
so this seems like a promising breakthrough but it’s still nowhere close to breaking encryption
Plus there is already some development into quantum proof algorithms, if my mind serves me correctly signal created the first one a few months ago
But can it run AI?
