HIGH PERFORMANCE COMPUTING
need to do the quantum chemistry for that to make an efficient process. And that kind of comes out around a million qubits. ‘Of course, there’s always more to the
story,’ stressed Earl. ‘We trade-off three things in a real-world application. One of them is the number of qubits, which is, of course, always important. There are also the error rates and coherence time. So if you have really good error rates, you can probably do fewer qubits, if you have long coherence times you can do with a few less qubits. So it’s a bit of a rough number. But that’s the kind of order of magnitude where quantum computers become interesting.’
Trapped ion quantum systems Universal Quantum is developing its quantum computers based on trapped ion technology. A simple explanation would be that ions are trapped and precisely controlled using electromagnetic fields. Each ion levitates above the surface of a silicon microchip. The idea behind trapped ion systems is the ions are relatively easy to control, as they are all precisely the same shape and size. ‘A trapped ion system is what you’d initially think of if you think of a quantum computer,’ said Earl. ‘The qubits you use are naturally quantum systems; an atom is the closest thing we can come to a single quantum bit. Whereas a superconducting qubit is kind of an analogy to a qubit. ‘The big benefit is that every qubit is identical because every atom is identical
to every other atom, the energy levels are very well defined, and especially with trapped ions, we can control the position and environment of those atoms very precisely,’ Earl continued. ‘By tuning electrode voltages, you can move the qubits around the surface of a chip very precisely; this means we have ultimate control over what these qubits are doing and how they’re interacting.’ Another important consideration that goes into developing these systems is the availability and price of the components used to control the qubits. For example, in the Universal Quantum system, the company uses lasers and microwaves to control the qubits. Earl noted it is important to focus on
developing systems with technology that is available today. ‘From the work – we’ve done at Sussex and the preliminary work – we’ve done at Universal Quantum already, we think we can build million qubit machines using technology that already
“We’re all very passionate about making real-world impactful systems as soon as possible. The quickest way to do that is using technology that already exists”
exists. We’re all very passionate about making real-world impactful systems as soon as possible. The quickest way to do that is using technology that already exists.’
Quantum accelerators Founded in 2019, Quantum Brilliance has a very different take on the development of quantum computing. The venture-backed company develops quantum computers using a diamond substrate to help boost the reliability of the qubits and increase coherence time. The goal of Quantum Brilliance is to enable mass deployment of quantum technology to propel industries to harness edge and supercomputing applications. The first generation of the company’s
technology has already been installed in Paswey Supercomputing Centre, which is exploring how this technology might be used alongside high performance computing (HPC) systems in the future. Mark Mattingley-Scott, managing
director, EMEA for Quantum Brilliance, explains why the company opted for this radically different quantum technology. ‘What diamond does is give you coherence for free. So you get qubits if you make qubits in diamond the right way. They maintain quantum coherence, even at room temperature. What it means is all the stuff you have to do with other quantum computing technologies, like keep it cold, or ensure it’s under a really high vacuum, or use precise lasers to get photons aligned, all those things fall away.’ This is an important distinction from other quantum systems, as it means the Quantum brilliance prototype systems can be smaller and more easily integrated with existing computing systems. They operate at room temperate and do not compex systems or advanced cooling. ‘I was actually at the Pawsey
Supercomputing Centre this afternoon and I saw our quantum computer,’ Mattingley-Scott said. ‘It’s a 6u device, so it’s a little bit higher than the standard 19-inch rack unit. That’s the first generation of machines. Inside this device is a small piece of diamond with the qubits on it. And some simple optoelectronics to interact with that – stuff you would find in a 5g antenna mast. We’re working on miniaturising that, and we’re pretty sure we can get down to a graphics-card-size accelerator within the next few years. ‘Once you’ve got something like a
graphics accelerator, then you’re in the same world as a normal Graphics Processing Unit (GPU) or Tensor Processing Unit (TPU),’ Mattingley-Scott added. ‘You can start to put these things
www.scientific-computing.com | @scwmagazine Summer 2022 Scientific Computing World 5
g
mapman/shutterstock
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38
