7 ms·

Here's a pdf of the article [1]. I'd love to hear a guru on quantum computation (maybe @michael_nielsen [2]) describe how quantum error correcting codes might i

Here's a pdf of the article [1]. I'd love to hear a guru on quantum computation (maybe @michael_nielsen [2]) describe how quantum error correcting codes might interface with this result. I thought the whole point of quantum computing was that it needed to be reversible, so if you add quantum ECC, can you still do it? I'm curious if the person who suggested or posted this to Hacker News knows if this is for sure cool, or if they just were dazzled (as I am) by the Q word. [1] https://physics.aps.org/featured-article-pdf/10.1103/PhysRevLett.115.190601 https://physics.aps.org/featured-article-pdf/10.1103/PhysRev... [2] https://twitter.com/michael_nielsen https://twitter.com/michael_nielsen

brudgers 11y ago
[IANAQCG] As long as quantum computations employ an arrow of time there is no way to escape the before and after semantics of the second law. My gut tells me that the arrow of time and simultaneity are a gestalt. Throwing out the arrow of time takes the baby with it, i.e. what would simultaneity get us without an after and before? Looking at the letter, once the experiment or equations requires initial conditions, there's no way to avoid an arrow of time. Change comes later. The arrow of time is a premise of the experimental method. For me, the headline parses out to a dog-bites-man "The Second Law of Thermodynamics" with "quantum" canceling itself out. As Chevy Chase would say, Charles De Gaulle is still dead.
q4h555qh5 11y ago
QECC takes "logical qubits on a larger space" to "corrected qubits on a smaller space". So while there is entropy production on the entire space (2nd law thermo), it is still possible to have entropy reduction locally. Eg: 0) "1 important qubit in some specific pure state" (entropy=0) + "4 ancillary qubits in some known but unimportant state"(entropy=0) are made using a 1to5 QECC into: 1) 1 logical qubit(made of 5-qubits) in a pure state (entropy=0) 2) Presence of noise: 5-qubits now in mixed-state (entropy=a>0) 3) Error correction on 5-qubits turning them into: 4) "1 important qubit in some specific pure state"(entropy=0) + "4 ancillary qubits in some now-random states" (entropy>=a>0)

brudgers 11y ago

[IANAQCG] As long as quantum computations employ an arrow of time there is no way to escape the before and after semantics of the second law. My gut tells me that the arrow of time and simultaneity are a gestalt. Throwing out the arrow of time takes the baby with it, i.e. what would simultaneity get us without an after and before? Looking at the letter, once the experiment or equations requires initial conditions, there's no way to avoid an arrow of time. Change comes later. The arrow of time is a premise of the experimental method. For me, the headline parses out to a dog-bites-man "The Second Law of Thermodynamics" with "quantum" canceling itself out. As Chevy Chase would say, Charles De Gaulle is still dead.

q4h555qh5 11y ago

QECC takes "logical qubits on a larger space" to "corrected qubits on a smaller space". So while there is entropy production on the entire space (2nd law thermo), it is still possible to have entropy reduction locally. Eg: 0) "1 important qubit in some specific pure state" (entropy=0) + "4 ancillary qubits in some known but unimportant state"(entropy=0) are made using a 1to5 QECC into: 1) 1 logical qubit(made of 5-qubits) in a pure state (entropy=0) 2) Presence of noise: 5-qubits now in mixed-state (entropy=a>0) 3) Error correction on 5-qubits turning them into: 4) "1 important qubit in some specific pure state"(entropy=0) + "4 ancillary qubits in some now-random states" (entropy>=a>0)