Getty Images

Nvidia QODA platform integrates quantum, classical computing

Nvidia's QODA platform bridges the chasm between quantum and classical environments. It could set the stage for quantum applications leveraging existing GPU and AI technologies.

Nvidia unveiled a hybrid platform that allows developers to create applications that work across classical and quantum systems to take advantage of performance-enhancing technologies including GPUs.

Designed to optimize the performance and scale of applications focused on scientific and quantum research, the Nvidia Quantum Optimized Device Architecture (QODA) platform also allows high-performance computing (HPC) developers to accelerate their existing classical applications for quantum environments.

"QODA is meant to be familiar for scientists working on scientific workloads, and also have interoperability with applications used today," said Tim Costa, director of HPC and quantum computing products at Nvidia. "Working with GPUs lets domain scientists incrementally add acceleration where it makes sense for existing classical applications."

The benefit of a unified programming environment like QODA, Costa said, is it not only removes performance bottlenecks, but it permits Nvidia to partner with leading algorithm developers and end users to bring their classical and quantum products together to solve a wider array of complex problems.

Tim Costa, director of HPC and quantum computing products, NvidiaTim Costa

QODA uses a kernel-based programming model and works with the C++ and Python programming languages. The environment also includes a compiler toolchain for hybrid systems and a standard library of quantum-based algorithmic primitives, according to Costa.

The environment is interoperable with a number of parallel programming techniques, which permits scientists to more easily move between running parts of third-party applications on classical resources as well as on quantum computing resources, Costa explained.

"We can allow domain scientists to leverage quantum acceleration and tightly couple it with the best of GPU supercomputing," Costa said.

The challenge in working across a hybrid classical-quantum environment is the processing lag time between the two environments, according to Paul Smith-Goodson, analyst in residence with Moor Insights & Strategy.

Getting an integrated classical-quantum environment to work is important if the industry is to achieve true quantum advantage. It's encouraging Nvidia is helping facilitate that because it's an important functionality that can't be overlooked.
Doug FinkeConsulting analyst, Inside Quantum Technology, and founder, Quantum Computing Report

Hybrid computation uses the quantum computer for part of the problem and a classical computer for other parts of the problem. This is common for algorithms such as the Quantum Approximate Optimization Algorithm (QAOA), where the problem is looped between a quantum computer doing quantum computations and a classical machine doing classical computing, Smith-Goodson explained.

"The back-and-forth swapping between the two technologies can significantly slow down the process," he said. "But if Nvidia's QAOA's open interface can streamline any part of the process, it will be helpful."

Getting an integrated classical-quantum system to work is an important step if the industry is to achieve true quantum advantage for commercially relevant applications, said Doug Finke, consulting analyst to Inside Quantum Technology and founder of the Quantum Computing Report.

"It's encouraging Nvidia is providing software that can help facilitate hybrid classical-quantum computing because it is an important functionality that can't be overlooked," he said.

One consultant said what appears to be missing from Nvidia's overall strategy is the specific role the cloud will play -- something every quantum computing player has laid out.

"To do a hybrid quantum environment, you need the cloud, and that piece of the strategy hasn't been made clear," said Frank Dzubeck, president of Communications Network Architects, Inc. "What they are trying to do with a unified programming environment and GPUs is interesting, but the cloud is what holds everything together. It's just the way things are done today."

When asked to provide further details about clouds, specifically popular cloud environments such as Microsoft Azure and AWS, Nvidia company officials said they would eventually support those environments.

Nvidia also announced during the Q2B 2022 conference held in Tokyo that it will work with quantum developers including IQM Quantum Computers, Pasqal, Quantinuum, Xanadu and Zapata Computing. It will also collaborate with the Lawrence Berkeley National Laboratory and Oak Ridge National Laboratory.

Providing an update on its CuQuantum SDK released last year, which helps accelerate quantum circuit simulations on GPU, the company said the offering is being used by "dozens of quantum organizations." QODA is designed to leverage CuQuantum's simulation environments.

The first beta versions of the new environment will be available to users by the end of this year, with general availability sometime in early 2023, the company said.

As Editor at Large with TechTarget's News Group, Ed Scannell is responsible for writing and reporting breaking news, news analysis and features focused on technology issues and trends affecting corporate IT professionals.

Next Steps

What Microsoft, Nvidia AI supercomputer partnership means

Nvidia DGX Quantum system blends CPUs, GPUs with CUDA

Dig Deeper on Data center hardware and strategy

SearchWindowsServer
Cloud Computing
Storage
Sustainability
and ESG
Close