The road to hybrid quantum-HPC data centers begins here

Start your journey into the future of high performance computing today with tools like NVIDIA cuQuantum.

It’s time to start building the hybrid quantum computers of the future.

Today, we have an irresistible motivation, a clear path, and the key components needed to build a hybrid quantum computer.

Quantum computing has the potential to solve some of today’s biggest challenges, advancing everything from drug discovery to weather forecasting. In short, quantum computing will play a huge role in the future of HPC.

Quantum simulation today

Inventing the future won’t be easy, but the tools we need to start the journey are here.

This is the first step toward enabling today’s supercomputers to simulate quantum computing jobs at a scale and performance level that is unattainable with existing relatively small and error-prone quantum systems.

Dozens of quantum organizations are already using the NVIDIA cuQuantum software development kit to accelerate their quantum circuit simulations on GPUs.

Recently, AWS announced the availability of cuQuantum in its Braket service. It also demonstrated on Braket how cuQuantum can achieve up to 900x speedup on quantum machine learning workloads.

cuQuantum is now available to accelerate computations on major quantum software frameworks, including Google’s qsim, IBM’s Qiskit Aer, Xanadu’s PennyLane, and Classiq’s Quantum Algorithm Design Platform, meaning users of these frameworks can access GPU acceleration without having to do any coding.

Quantum-driven drug discovery

Today, Menten AI is beginning to use cuQuantum to support its quantum work.

The Bay Area drug discovery startup will use cuQuantum’s Tensor Networks library to simulate protein interactions and optimize new drug molecules. The goal is to use quantum computing’s potential to accelerate drug design, a field that, like chemistry, is widely considered to benefit first from quantum acceleration.

Specifically, Menten AI is developing a suite of quantum computing algorithms, including quantum machine learning, to solve computationally intensive problems in therapeutic design.

“While quantum computing hardware capable of running these algorithms is still under development, classical computing tools such as NVIDIA cuQuantum are essential to advancing the development of quantum algorithms,” said Alexey Galda, chief scientist at Menten AI.

Building a quantum link

As quantum systems develop, the next big leap is toward hybrid systems: quantum and classical computers working together. Researchers hope that these system-level quantum processors, or QPUs, will become powerful new accelerators.

Therefore, an important task ahead is to bridge the classical and quantum systems into a hybrid quantum computer. This task consists of two main parts.

First, we need to establish a fast, low-latency connection between the GPU and the QPU. This will allow the hybrid system to use the GPU for the tasks it traditionally excels at, such as circuit optimization, calibration, and error correction.

GPUs can shorten the execution time of these steps and significantly reduce the communication latency between classical and quantum computers, which is the main bottleneck facing hybrid quantum operations today.

Second, the industry needs a unified programming model with efficient and easy-to-use tools. Our experience in HPC and AI has taught us and our users the value of a solid software stack.

The right tool for the job

Currently, to program QPUs, researchers can only use the quantum equivalent of low-level assembly code, which is inaccessible to scientists who are not experts in quantum computing. In addition, developers lack a unified programming model and compiler toolchain, so they cannot run their work on any QPU.

This needs to change, and we believe it will. In a March blog post, we discussed some of our initial work toward a better programming model.

In order to efficiently find ways to accelerate work on quantum computers, scientists need to easily port parts of their HPC applications first to a simulated QPU and then to a real QPU. This process requires a compiler that allows scientists to work efficiently in a familiar way.

By bringing together GPU-accelerated simulation tools, programming models, and compiler toolchains, HPC researchers can begin building the hybrid quantum data centers of the future.

Getting Started

To some, quantum computing may sound like science fiction , decades in the future, but researchers are building more and larger quantum systems every year.

NVIDIA is fully engaged in this effort and invites you to join us to begin building the hybrid quantum systems of the future today.

To learn more, you can watch the GTC session and attend the ISC tutorial on the topic. For a deeper dive into how GPUs are used today, check out our State Vector and Tensor Networks libraries.