The University of Sydney team, backed by Microsoft and led by the world-renowned Australian quantum physicist Professor David Reilly, says it has now figured out how to let qubits remain at millikelvin temperatures, and still run sufficient control wires to them, to tell the qubits what to do and then read out the results, all from a computer operating at the relatively scorching room-temperature required by human operators.
The invention is made up of two main components: a chip that sits alongside the qubits on the coldest, bottom shelf of the “dilution” refrigerator, controlling the qubits and reading their output; and a second chip, joined to the first by just two wires, that takes the load off the first chip and that sits one shelf higher in the fridge, where it can afford to give off a little more heat without upsetting the delicate qubits.
(In order to not interfere with the strange quantum properties of the qubits, the coldest chip in Gooseberry only draws power measured in the “nanowatts”, or billionths of a watt. By way of comparison, the main Intel processor in a PC might consume 30 watts.)
The second Gooseberry chip then connects to the world outside the fridge with just three wires, carrying digital signals to and from the classical computer that controls the quantum computer.
It’s this small number of wires leaving the fridge – just three, regardless of how many thousands or millions of qubits there are inside the fridge – that’s the trick.
“The idea of simply having a lot of cables that run from 100 millikelvin in a dilution refrigerator up to room temperature – well, if you’ve seen pictures of the Google refrigerator that did the quantum supremacy experiment, you’d appreciate that that is really not an approach that can scale to millions of qubits,” said Professor Andrew Dzurak, one of the world’s top quantum computer scientists who works at the University of NSW, who was not involved in Professor Reilly’s project.
Professor Dzurak said the Sydney University work on Gooseberry was “very impressive”, all the more so because it could be applied to many different types of quantum computers, not just to computers using the “topological” qubits that Microsoft had hired Professor Reilly to work out how to control.
However, Professor Dzurak said it still remained to be seen whether Gooseberry could scale up to the millions of qubits that a quantum computer would need to perform real-world computations, and still maintain the low temperature required.
If it could, the technology might even be something he would use for his own work at UNSW, which has been pursuing a different type of qubit to Microsoft’s, but which also had the problem of scaling up to millions of qubits without adding too much heat.
Professor Reilly told the Australian Financial Review that, although his experiments outlined in Nature only had around 64 outputs for controlling qubits, it was already capable of controlling thousands of qubits, all from three wires leading to the outside world.
“We didn’t just keep on adding output cells for the sake of it, without having thousands of qubits to test. So we just stopped at 64 or so,” he said.
Although Gooseberry essentially meant he had now made the breakthrough that Microsoft had hired him for (and invested tens of millions of dollars in the University of Sydney in the process), there was still much work to be done, Professor Reilly said.
“We’re going to improve this. We’re going to build out more complexity that allows us to (control) even more qubit platforms; to improve the fidelity of the control; and to scale to even larger numbers.
“Why not start thinking about billions of qubits? The more qubits we can control, the better,” he said.
This is a syndicated post. Read the original post at Source link .