BOFFINS CLAIM they are close to making accurate simulations of reality on the molecular scale using quantum computers.
James Whitfield, a quantum information chemist at Harvard University has a cunning plan to simulate molecules on computers to better understand how they might react. This might, for example, give researchers a good idea about how a drug might behave in the human body.
So far simulating complex molecules using modern supercomputers falls short because increasing the number of atoms they have to analyze takes too long. If you try to look at more than four or five atoms in a chemical reaction, or even a moderately complex molecule, it quickly becomes a computational headache.
Whitfield said that at best a regular computer only can get a rough idea of how these systems work. A quantum computer, which apparently is powered by a cat which is either there or not there, enables scientists to carry out two calculations simultaneously.
This is because it uses the computing power of both potential cats using slide rules and pooling their results. At least that was the way we recall it all having been explained at school.
Alán Aspuru-Guzik, another Harvard quantum chemist, said that if you wanted to simulate quantum systems it is probably a much cleverer idea to use another quantum system to do it.
While Aspuru-Guzik, Whitfield and their colleagues provided the software and performed key calculations, their collaborators in Australia assembled the hardware and ran the experiments.
Using a two-qubit computer, they simulated the smallest molecular system, the hydrogen molecule, and calculated its energy in terms of how it might react with other molecules.
After running the simulation 20 times in a row, with each cycle working off the data from the last one, apparently the slide rules wielding cats managed to come up with a pretty precise value.
The next thing to do is develop a more powerful quantum computers with shedloads more qubits, and more cats, we guess.
They think that using a 2,000 qubit quantum computer would give them roughly enough computation power to simulate cholesterol binding with a protein.
That sounds to us like it's going to take a lot of cats, even if half of them would not actually be there. Unless we have got it wrong, which we suppose we might have. µ
Chinese firm beats rival Samsung to the punch... er, hole
Getting botter all the time
It's the best of the rest from Google's week
Just like we promised ourselves we wouldn't do again