- QPUs could run AI inference quicker and cheaper than typical {hardware} ever may
- Hybrid nodes combining CPUs, Nvidia GPUs, and diamond QPUs may change how we construct quantum software program
- From protection to finance, Quantum Brilliance is betting on diamond chips to drive adoption
Diamonds have emerged as a important materials within the growth of quantum applied sciences on account of their distinctive atomic properties, and Quantum Brilliance, an organization primarily based in Germany and Australia, has outlined an formidable plan to develop moveable quantum computer systems utilizing diamond-based quantum processing items (QPUs).
These units are being designed to function at room temperature and will finally be built-in alongside GPUs and high-end CPUs in servers or autos.
However whereas the corporate’s imaginative and prescient guarantees a future the place quantum computing is as seamless as plugging in a GPU for AI inference, a number of technical and industrial hurdles stay.
Rethinking quantum computing with diamonds
Over the previous decade, researchers have more and more targeted on engineering high-purity artificial diamonds to reduce interference from impurities.
Notably, a 2022 collaboration between a Japanese jewellery agency and tutorial researchers led to a brand new methodology for producing ultra-pure 2-inch diamond wafers.
In 2023, Amazon joined the hassle by way of its Middle for Quantum Networking, partnering with De Beers’ Ingredient Six to develop lab-made diamonds to be used in quantum communication techniques.
Now, Quantum Brilliance goals to make the most of nitrogen vacancies in diamond to create qubits, providing a extra compact and power-efficient various to cryogenic quantum techniques.
“We do have a roadmap to fault tolerance, however we’re not worrying about that in the meanwhile,” mentioned Andrew Dunn, COO of Quantum Brilliance.
“Folks consider hundreds of thousands of qubits, however that shall be very costly and energy hungry. I feel getting an understanding of getting 100 qubits in a automotive cheaply and easily – the use circumstances are very totally different.”
This alerts a departure from the prevailing pattern in quantum computing, which focuses on constructing techniques with hundreds of thousands of qubits.
The corporate is as a substitute concentrating on cheap and sensible use circumstances, notably in purposes akin to AI inference and sparse information processing.
Quantum Brilliance is already collaborating with analysis establishments just like the Fraunhofer Institute for Utilized Strong State Physics (IAF).
IAF is presently evaluating the corporate’s second-generation Quantum Growth Package, QB-QDK2.0, which integrates classical processors like Nvidia GPUs and CPUs with the QPU in a single field.
In parallel, Oak Ridge Nationwide Laboratory within the US has acquired three techniques to review scalability and parallel processing for purposes like molecular modeling.
“The explanation they’re shopping for three techniques is that they wish to examine parallelisation of techniques,” Dunn added.
Quantum Brilliance can be working carefully with imec to combine diamond processes into normal chip manufacturing.
Past computation, the corporate sees potential in quantum sensing, and the expertise might also be repurposed for protection and industrial sensors.
Finally, the corporate needs quantum computing to grow to be as unusual as another chip in a server.
“Personally, I wish to make quantum actually boring and invisible, simply one other chip doing its job,” mentioned Dunn.
By way of eeNewsEurope
