Artificial Intelligence Giant NVIDIA creates a new Platform for Hybrid Quantum-Classical Computing: QODA

Artificial Intelligence Giant Nvidia Creates A New Platform For Hybrid Quantum-Classical Computing: Qoda

Nvidia has been steadily moving further into the quantum space. We have seen previous efforts of the semiconductor move into quantum simulation. This latest initiative is aimed at capturing more of the the quantum workflow or rather the quantum-classical workflow. For developers who want to do algorithm research on hybrid applications for future quantum advantage, bridging technology is needed to enable dynamic workflows across disparate architectures: welcome QODA, which NVIDIA claim is the worlds first platform for hybrid quantum classical computing.

Simulating Quantum

Previously Nvidia announced cuQuantum, a software development kit (SDK) for accelerating simulations of quantum computing workflows. Researchers develop Simulating quantum circuits using cuQuantum on GPUs enables algorithms research with performance and scale far beyond what can be achieved on quantum processing units (QPUs) today. Before Quantum Computers are ready for prime time, researchers must simulate on devices with the appropriate number of qubits and as the number of qubits increases ever more powerful classical or conventional computers must be used to hit every higher numbers of qubits.

Qoda Is Both A Specification And A Compiler Delivering A Unified Programming Model Designed For Quantum Processors (Either Actual Or Emulated).  Devices Such As Cpus, Gpus, And Qpus Can All Work Together In A Smart Coherent Way With The Ease Of Frameworks That Developers Know And Love Such As Cuda.
QODA is both a specification and a compiler delivering a unified programming model designed for quantum processors (either actual or emulated). Devices such as CPUs, GPUs, and QPUs can all work together in a smart coherent way with the ease of frameworks that developers know and love such as CUDA.

QODA for High Performance Computing

Just as NVIDIA has popularised and standardised much of parallel computing on GPU’s such as CUDA. The graphics giant behind some of the most popular graphics cards has announced the QODA programming model specification and corresponding NVQ++ compiler platform. As expected QODA will be interoperable with parallel computing framework CUDA (developed by NVIDIA) and OpenMP and OpenACC.

QODA will enable an efficient programming environment for accelerating hybrid algorithm research and development. Scaling can happen right the way from laptops to massive multi GPU systems. It will be backend agnostic meaning that it can work with a whole series of different qubit types and implementations.

QODA from NVIDIA is a first-of-its-kind platform for hybrid quantum-classical computers, enabling integration and programming of quantum processing units (QPUs), GPUs, and CPUs in one system.

Similar usage to CUDA,

QODA provides a CUDA-like kernel-based programming approach. Like other frameworks quantum device code can be represented with standalone function objects or lambdas annotated with __qpu__ to indicate to the compiler that this code is executed on a quantum device.

  • Open. Connects to any type of QPU backend, allowing accessibility to all users
  • High Performing. 287X speedup in end-to-end Variational Quantum Eigensolver (VQE) performance with 20 qubits and dramatically improved scaling compared to Pythonic frameworks
  • Easily Integrated. Interoperates with modern GPU-accelerated applications
  • Productive. Streamlines hybrid quantum-classical development with a unified environment, improving productivity and scalability in quantum algorithm research

There is an early release program available too.