Nvidia launches quantum computing platform

The goal is to make it easier to program in quantum computing, which is very different from standard computing.

Quantum computing  >  A quantum processor radiates power.

Nvidia, the darling of high performance computing (HPC), is bringing new attention to quantum computing. 

The company has launched its Nvidia Quantum Optimized Device Architecture, or QODA. This hybrid platform is designed to make quantum computing more accessible by enabling programming of both quantum applications and classical applications in a single, consolidated environment. According to Nvidia, it's aimed at speeding breakthroughs in quantum research and development across AI, HPC, health, finance and other disciplines.

The aim is to make QODA be for quantum computing just as CUDA is for GPU computing – an industry standard. (CUDA is a C-like code for writing specialized HPC and AI applications that run on Nvidia GPUs.)

Nvidia said HPC and AI developers can use QODA to add quantum computing to existing applications, leveraging both quantum processors as well as simulated future quantum machines using Nvidia DGX systems and Nvidia GPUs available in scientific supercomputing centers and public clouds.

This is not Nvidia’s first dance in the quantum computing space. About a year ago, the company released cuQuantum, a software development kit (SDK) for accelerating quantum workflows using the Tensor Cores in its GPUs along with various libraries and tools optimized for such jobs as quantum circuit simulations.

In announcing the new architecture, Nvidia has announced collaborations with a host of quantum computing firms, most with a Q in their name: hardware vendors IQM Quantum Computers, Pasqal, Quantinuum, Quantum Brilliance and Xanadu; software providers QC Ware and Zapata Computing; and supercomputing centers Forschungszentrum Jülich, Lawrence Berkeley National Laboratory and Oak Ridge National Laboratory, which is interesting because ORNL is an all-AMD shop.

What is quantum computing?

The processes have advanced, but the basic structure of computing has not changed since it was invented. Data is represented at its most basic state as bits, 0 or 1. Quantum computing uses something called a quibit, which can represent a 0, a 1, or any proportion of 0 and 1 in superposition of both states. A quibit can be 1/4 0 and 3/4 1, for example.

This means multiple things. First, much greater speed. Quantum computing can process data at up to 1,000 times faster than standard binary computers. You’re not going to use a quantum computer to run Microsoft Excel. You’re going to use a quantum computer to do weather simulations and drug testing.

Second, quantum computing has no compatibility with current applications. You’re not just going to rewrite or recompile an application on a quantum computer. You have to write the whole thing over from scratch to completely take advantage of the new architecture. No one is going to find that appealing, and that’s what Nvidia is trying to address.

Fortunately, it’s not an either/or situation. Applications that will be accelerated by quantum processing units (QPU) will be hybrid workloads that leverage a standard supercomputing architecture for large parts of an application, while the most critical parts are accelerated by a quantum system.


Copyright © 2022 IDG Communications, Inc.

The 10 most powerful companies in enterprise networking 2022