Nvidia Corp. today detailed three initiatives in which researchers will use its chips to support scientific discovery.

One initiative will see the University of Bristol deploy a supercomputer powered by Nvidia’s Grace CPU Superchip processors. The two other projects focus on quantum computing. The chipmaker detailed all three projects at the ISC 2023 conference taking place this week in Hamburg.

Energy-efficient supercomputing

In the U.K., the University of Bristol is leading an initiative to build a new supercomputer optimized for research workloads. The supercomputer will be known as Isambard 3. It’s set to be built by Hewlett Packard Enterprise Co. using Nvidia silicon.

According to the chipmaker, Isambard 3 will feature 384 of its Grace CPU Superchip processors. Each such processor includes 144 central processing unit cores based on an Arm Ltd. design. The cores support SVE, an Arm instruction set technology that allows CPUs to run artificial intelligence workloads more efficiently.

Nvidia expects that the Isambard 3 will provide maximum speeds of 2.7 petaflops when performing calculations involving FP64 data. One petaflop equals 1 quadrillion computing operations per second. FP64, in turn, is a data format commonly used by supercomputers to organize the information they process.

Scientists plan to use Isambard 3 for a range of tasks. According to Nvidia, the system will support research initiatives in fields such as AI, life sciences, medicine, biotechnology and astrophysics. It will be capable of simulating both large, complex structures such as wind turbines and molecular-level biological mechanisms.

Isambard 3 is scheduled to come online in 2024. Once operational, the system is expected to consume less than 270 kilowatts of power, which will make it one of the most energy-efficient supercomputers in Europe. Nvidia estimates that Isambard 3 is six times more energy-efficient than the University of Bristol’s previous-generation supercomputer.

“As climate change becomes an increasingly existential problem, it’s vital for computing to embrace energy-efficient technologies,” said Ian Buck, the vice president of hyperscale and high-performance computing at Nvidia. “NVIDIA is working alongside the Arm Neoverse ecosystem to provide a path forward for the creation of more energy-efficient supercomputing centers, driving important breakthroughs in scientific and industrial research.”

New quantum computing initiatives 

At ISC 2023 today, Nvidia also detailed how its technology is helping to support two quantum computing initiatives in the U.K. and Germany.

The first initiative is led by Rolls-Royce Holdings plc in partnership with Classiq Technologies Inc., a quantum computing startup. The project’s goal is to advance Rolls-Royce’s efforts to develop more efficient jet engines. The company hopes to use quantum computers to ease its engineers’ work.

Working with Classiq, Rolls-Royce used a set of Nvidia A100 graphics cards to develop and simulate a quantum circuit. That’s a type of specialized algorithm that could in the future run on a quantum computer. Rolls-Royce says that its algorithm has 10 million “layers” and 39 qubits.

According to Nvidia, the quantum circuit is designed to perform calculations related to computational fluid dynamics, a scientific field focusing on simulating complex physical phenomena. Rolls-Royce’s quantum circuit is touted as the largest of its kind developed specifically for computational fluid dynamics.

“Designing jet engines, which are one of the most complicated devices on earth, is expensive and computationally challenging,” Buck said. “NVIDIA’s quantum computing platform gives Rolls-Royce a potential path to tackle these problems head on while accelerating its research and future development of more efficient jet engines.”

The second quantum computing initiative Nvidia detailed today is a collaboration with FZJ, a research institution in Germany. FZJ is partnering with Nvidia to build a new lab that will feature what the companies describe as a classical-quantum supercomputer. The system is set to include both classical and quantum computing elements.

Nvidia detailed that the machine will be developed by ParTec AG, a Munich-based company. FZJ researchers, in turn, will use Nvidia’s cuQuantum and CUDA Quantum software tools to simulate quantum processors on the system. FZJ already operates a classical supercomputer that features 3,744 of Nvidia’s A100 graphics cards. 

Image: Nvidia