Quantum chip lab opens in London

Quantum chip lab opens in London

Interviews |
By Nick Flaherty

Quantum Motion has opened a laboratory in London as the next stage in the development of a CMOS chip for quantum computers.

“We came from being in a couple of universities with contract R&D work and moving here allows integrated circuit and quantum specialists to have direct connection with the foundry,” said James Palles-Dimmock,  chief operating officer of Quantum Motion talking to eeNews Europe.

The company has developed a silicon qubit tile for quantum computers that uses existing silicon foundry technologies based on research at the University College London and the University of Oxford.

“Silicon qubits work at 1K and above and that’s a lot higher than a lot of superconducting qubits but we go all the way down to 10mK as that helps the R&D process,” he said. The lab has dilution fridges for testing system at those ultra-low temperatures but the practical applicant will fit in a 19in ack in a data centre, says Palles-Dimmock

“The way our qubits work with something that is a lot like a gate, but at sub-threshold voltages and suck up a single electron under the gate and that’s why you need the low temperatures,” he said. “You then add the digital to analog converters DACs etc and put hundreds and millions of them on a chip,” he said. “It works as a transistor – it looks in size and shape like a conventional transistor and uses the same processes and you can even monolithically  integrate them,” he said.

Related quantum articles

“You have two transistors close together and you can move electrons between them – if you can’t move them the spins are aligned, if you can move them then they are not. You then perform a simple charge measurement so you have a spin to charge transfer – it’s a neat way to do it,” he said.

“This are small signals so we need amplification and we match these with large inductances to measure shifts in resonant frequencies. “We do the design work, take to a foundry and get it back and test, and we have used more than one foundry,” he said. These include CEA-Leti (below) and other tier one foundries.

“It doesn’t need to be bleeding edge technology, anything below 100nm that can form good quality quantum dots,” he said. “We break a few small rules which we can have waivers for, and Cadence have supported us really well, giving us recommendations on the design tools. It’s about how to control and interface them, it’s about the analogue tweaks you needs.”

The company is chaired by Alberto Sangiovanni-Vincentelli, one of the founders of Cadence Design Systems and Synopsys, and this has been a significant help says Palles-Dimmock.

“We’ve got really good backing from investors and grants with $20m and we are a team of 33,” he said, “What’s helped as well is having Alberto as chair validates the technology and brings in additional people.”

Quantum hardware

The company focus is on the hardware, developing and validating a four qubit tile that can be replicated on a single chip with all the supporting control and readout circuits and a conventional computer as the controller. All of this, with the cooling, would fit in a 19in rack in a data centre

“We developed a tile with four qubits, one on each corner and one in the middle for error correction and that could go up to millions of qubits,” he said. “For a fault tolerant architecture you need 1000 physical qubits to give one logical qubit so a million gives you 1000 logical qubits. Applications need 300 to 1000 perfect qubits and that’s well within our roadmap. The smallest sensible sized 50 to 100 cubits and we can develop that by linking up multiple tiles.”

“We are looking to put this in a data centre with partners and the benefit is we can integrate into the systems as a hardware provider,” he said.

The lab is a key step in the validation and testing of the technology.

“We have proven high quality qubits can exist in silicon but we haven’t shown that in an industrial process,” he said, “We have test chips back from foundry.”

“We see the biggest challenge as the hardware,” he said, rather than high level quantum operating systems. “I don’t believe that a quantum computer OS is a sensible thing. We have very small qubit machines and we still will do this for a while and you are doing [the calculations] directly on the chip.

The validation of the technology is also key to the next round of funding planned for the near future to accelerate the process of integration and work with the foundries. “Our Series A round was three time oversubscribed and there’s appetite for this and I think that’s because we are unique. If you are looking for standard silicon for qubits it’s us or Intel and that is attractive to investors.

Other quantum articles

Other articles on eeNews Europe


Linked Articles