NVIDIA has launched a unified computing platform for quantum research and development across AI, HPC, health, finance and other disciplines.
Called NVIDIA Quantum Optimized Device Architecture, or QODA, this is the successor to CUDA – Compute Unified Device Architecture.
NVIDIA’s ambition is to solve the long-standing issue in the quantum world: to build quantum interoperability, create better accessibility for traditional DevOps roles, all while preserving its open nature.
As a hybrid platform, QODA uses standard Python, Java, and C++ programs for deployment while creating a quantum-driven kernel architecture. Moreover, QODA’s NV++ compiler is backend-agnostic and single-sourced. It supports both quantum and classical computing translations, allowing developers to switch between the two as and when a problem statement comes into the picture.
QODA can run on both GPUs and physical quantum processor units (QPUs). The idea is to increase a program’s portability across physical and emulated environments, along with building a standard library of quantum primitives in parallel.
NVIDIA wants to provide hardware diversity through qubit “modalities”, including diamond processors, superconducting, photonics, neutral atom and trapped ions. QODA’s qubits in transit could race through complex maths in minutes, something a classical computer will take hundreds of years to compile, let alone finish.
The company has created QODA’s architecture by combining DGX Quantum and cuQuantum SDK, helping developers to locate quantum circuits on GPUs while offering integrations with computing frameworks like Pennylane, Cirq and Qiskit. QODA is compatible with NVIDIA’s other computing platforms, including CUDA, OpenMP and OpenACC.
- Under pressure – NVIDIA Calls Off £29bn Arm Acquisition – read the story here
QODA comes at a time when NVIDIA’s GPU technology is doing well within the HPC market. Precedence Research says the HPC market size was valued at $39.86 billion (£33.6 billion) in 2021. Based on that research, HPE, IBM, AMD and Microsoft are also key players in the North American market.
The limited capacity of GPUs as quantum hardware cannot make them a satisfactory medium for quantum circuit emulation – one of the reasons behind QODA’s ideation. As of now, QODA is actively looking for developers with an early expression of interest. NVIDIA has asked interested developers to apply for a potential role via its developer site.
NVIDIA is yet to comment on QODA’s cloud integration capabilities – something that the quantum industry has realised as the ‘minimum benchmark’. However, the enterprise has announced support for AWS, Azure and other popular cloud environments, in the later project roadmaps.
At the official launch, NVIDIA announced QODA collaborations with quantum hardware providers IQM Quantum Computers, Pasqal, Quantinuum, Quantum Brilliance and Xanadu; software providers QC Ware and Zapata Computing; and supercomputing centeres Forschungszentrum Jülich, Lawrence Berkeley National Laboratory and Oak Ridge National Laboratory.
There have been plenty of other developments within the field of quantum computing. A few examples are below.
In May, trade association techUK was advocating for the propagation and commercialisation of quantum technologies and industry across the UK.
While in March, HSBC said it will work with IBM as it seeks to explore applications for quantum computing in financial services.
Also in March, the NATO Cyber Security Centre revealed it had been experimenting with post-quantum encryption to ensure secure communications flows are protected against attacks that use quantum computing.
Image courtesy of IBM Quantum Computers.