Intel is taking a slow and steady approach to quantum computing. Competitors like Google may be聽racing to achieve so-called quantum supremacy, in which a quantum computer outperforms an ordinary one. But Intel鈥檚 James Clarke has bigger ideas. He leads the firm鈥檚 quantum computing research team, and says it is looking past near-term goals in聽order to be the first to make a device with a million qubits, or quantum bits 鈥� enough to have a聽real impact on the world.
How are you making a quantum computer?
We have spin qubits in silicon, which are like single electron transistors. We make billions of transistors today, so the thought is, if we can turn these into qubits聽鈥� which is non-trivial聽鈥� they would have some key [scaling] advantages over the superconducting qubit [Google鈥檚 preferred approach]. We aren鈥檛 putting all our eggs in one basket, and the superconducting qubits are a little further along. We have a聽49-qubit chip of those.
Why go for the long term?
What do I define as long-term success? Something that would change your life or mine聽鈥� maybe a new drug. For that, you need very good qubits and you need a聽lot of them.
Advertisement
What is Intel鈥檚 big goal?
Right now, the whole community is at, let鈥檚 say, tens of qubits, maybe 50. And maybe with brute force you could make a chip bigger and bigger and get to about聽1000. With 1000, you can probably do some interesting things. You can probably even find a couple of applications where you are doing a little bit better than a supercomputer.
My goal would be to find a way to get a million or a billion qubits together on a chip. That keeps me up at night. At Intel, we are less worried about the answers we will聽get from 50 qubits and more about how we will get to a million. We want to be the first to do that.
When will that be possible?
We are saying roughly 10 years. Some of the other companies in this field are saying something much nearer term. But if you look聽at the history of the evolution聽in microelectronics, it聽actually happened on a little bit聽longer scale than that. The first聽silicon transistor was in 1954, the聽first integrated circuit was in 1958 and the first microprocessor was in 1970.
Will I use a quantum computer in my lifetime?
The Cray-1 supercomputer came out in the mid-70s. I doubt anybody then would have said, 鈥淗ey, I bet 40 years from now, we are going to have these in our back pockets for listening to music and watching television.鈥� It is hard to know where we will be in 30 to 40 years, so I wouldn鈥檛 rule it out.
On a more practical level, the first quantum computers will be hooked up to a supercomputer. Chances are, it will take a team of聽experts who are familiar with programming the supercomputer and programming a quantum computer to get the information in or out. We need to develop the workforce to be able to do that. It doesn鈥檛 really exist yet. Quantum computer programmers are few and far between.
