Nvidia expands NVQLink quantum interlink as it preps for future quantum processor connectivity
At this week’s Supercomputing Conference, SC25 in St. Louis, Missouri Nvidia highlighted how it expects to accelerate computing to quantum processors in the future.
Nvidia only in October introduced NVQLink interconnect to let lets GPUs and quantum chips to each other. The package is built on the NVLink interconnect for direct CPU-to-GPU communications. Nvidia might be getting ahead of itself, since there are no quantum processors on the market nor will there be for some time. There cloud-based services from D Wave, Rigetti, IONQ, AWS and others.
“In the future, every supercomputer will draw on quantum process to expand the problems it can compute, and every quantum processor will rely on a supercomputer to run correctly,” said Dion Harris, senior director of HPC and AI infrastructure solutions at Nvidia on a pre-conference call. “GPUs act as the brain that orchestrates the quantum hardware and perform the heavy computations that quantum processors cannot.”
[ Related: More Nvidia news and insights ]
Paired with NVQLink is CUDA-Q, which provides a programming model where quantum processors (QPU), GPUs, and CPUs all work together in the same application. This is necessary because most useful quantum algorithms will rely on traditional computing for tasks like control, optimization, and error correction.
The first quantum computing company to support NVQLink will be Quantinuum, with its Helios quantum processor. Nvidia says that by uniting quantum processors with Nvidia accelerated computing, NVQLink’s open system architecture overcomes control and error-correction challenges and enables the development of hybrid quantum-classical applications. It delivers 40 petaflops of AI performance at FP4 precision with a GPU-QPU throughput of 400 Gb/s and a latency of less than four microseconds.
Overall, Nvidia has an impressive lineup of science-oriented firms and institutes lining up to use NVQLink. They include science and technology institutes in Japan, Korea, Taiwan, Singapore and Australia in the APAC region.
Europe and the Middle East are also lining up behind quantum computing research, with supercomputing and quantum technology centers in Italy, Denmark, the Czech Republic, Germany, Poland, and Saudi Arabia.
In the U.S., a number of National Laboratories are signed up to use NVQLink, including labs at Los Alamos, MIT, Sandia, Berkeley, and of course Oak Ridge National Laboratory.
More Nvidia news:
- Nvidia highlights considerable science-based supercomputing efforts
- Nvidia’s first exascale system is the 4th fastest supercomputer in the world
- Nvidia highlights considerable science-based supercomputing efforts
- Next-generation HPE supercomputer offers a mix of Nvidia and AMD silicon
- Cisco, Nvidia strengthen AI ties with new data center switch, reference architectures
- Nvidia looks to power AI factory networks
- Should enterprise developers care about Nvidia?
