- Quantum Optimized Device Architecture (QODA), a new platform from Nvidia, will establish a single, open environment for both conventional computers and quantum processors.
- By the end of the year, Nvidia hopes to have a closed beta of QODA running. The platform is going to be cost-free when it launches.
The advent of quantum computing is leading to significant advancements in various fields, including investment, engineering for electric vehicles, and pharmaceuticals.
But in reality, most applications that drive these advancements won’t entirely depend on quantum computing. They’ll probably employ hybrid quantum-classical computing, with specific algorithms running on classical computer infrastructure and others optimized for quantum computing.
However, a piece of the puzzle is still missing: A method for most High-Performance Computing (HPC) developers to use quantum computers to speed up their current applications.
Quantum Optimized Device Architecture (QODA), a new platform being introduced by Nvidia, promises to establish a single, open environment for both conventional computers and quantum processors to solve this issue.
According to Nvidia, developers can use open and interoperable standards across different software and hardware platforms, and hybrid quantum-classical computing will be within reach.
Tim Costa, Nvidia’s director of HPC and quantum computing products, believes that the platform would prompt a “revolution in the accessibility of hybrid quantum-classical computing.”
Thanks to QODA’s programming model and compiler toolchain, modern scientific computing applications can be quantum accelerated. This means that the toolchain and programming model was created to integrate into and interoperate with the tools, methods, and languages that current scientific developers already use in their applications.
Nvidia hopes to have a closed beta of QODA running by the end of this year. The platform is going to be cost-free when it is launched.
Nvidia does not create quantum processors; it creates the traditional GPU supercomputing infrastructure that will coexist in a hybrid system with quantum processors. As a result, the QODA platform is an open platform. Nvidia is building it in conjunction with the top manufacturers of quantum computing hardware, software, and research organizations.
At the launch, Nvidia will have five leading quantum computing hardware providers from different qubit modalities supported by QODA. These five include Pasqal, Xanadu, IQM Quantum Computers, Quantum Brilliance, and Quantinuum. Software companies like QC Ware and Zapata and supercomputing centers at Oak Ridge National Laboratory, Forschungszentrum Jülich, and Lawrence Berkeley National Laboratory are a few launch partners.
Costa compared the development of QODA by Nvidia to that of CUDA, the company’s parallel computing platform and programming model that enables programmers to use GPUs to speed up their computationally expensive applications.
“It’s really part of our DNA as a company to improve the accessibility of disruptive technologies,” Costa said. “We did it with GPU computing, and now we’re looking at the same problem in hybrid quantum-classical computing.”