Quantum computing guarantees to shorten the trail to fixing among the world’s greatest computational challenges, from scaling in-silico drug design to optimizing in any other case impossibly advanced, large-scale logistics issues.
Integrating quantum {hardware} into state-of-the-art AI supercomputers — forming accelerated quantum supercomputers — helps velocity the scaling of at the moment’s quantum processors into useful units for fixing these advanced challenges.
On the COMPUTEX commerce present, NVIDIA underscored how its work with companions throughout the Taiwan supercomputing ecosystem is advancing quantum computing towards accelerated quantum supercomputers.
Main {hardware} builders are working with NVIDIA to equip quantum researchers with instruments to make vital contributions within the area.
Atlantic Quantum, the College of Edinburgh,, the College of Oxford, Quantum Circuits Inc., QuEra Computing and Yale College anticipate receiving NVIDIA Grace Hopper Superchips from Supermicro to discover and refine the intersections between AI supercomputing and quantum computing.
Compal introduced its CGA-QX platform, constructed utilizing the NVIDIA CUDA-Q platform, to speed up the simulation of quantum optimization issues. The platform has been adopted by the Taiwanese Nationwide Science and Expertise Council and made obtainable to researchers from universities throughout Taiwan.
Quanta has employed NVIDIA CUDA-Q to experiment with bodily quantum {hardware}, utilizing the platform’s state vector simulations to confirm and validate current quantum processors. This permits Quanta to know the small print of noise in its methods and assess how properly they work to be used instances of curiosity.
NVIDIA can be working with supercomputing facilities to advance accelerated quantum supercomputing.
Taiwan’s Nationwide Heart for Excessive-Efficiency Computing (NCHC) introduced a brand new supercomputer for quantum analysis. Constructed by ASUS, the AI supercomputer at NCHC will function NVIDIA HGX H200 methods with over 1,700 GPUs, two NVIDIA GB200 NVL72 rack-scale methods and an NVIDIA HGX B300 system constructed on the next-generation NVIDIA Blackwell Extremely platform — interconnected by NVIDIA Quantum InfiniBand networking. Introduced at the moment at COMPUTEX, it’s anticipated to go reside later this yr.
NCHC is enabling over 20 firms working in quantum computing — collaborating underneath what’s generally known as the Nationwide Quantum Workforce. NVIDIA CUDA-Q is getting used on the middle to discover quantum options for purposes starting from machine studying to chemistry.
In Japan, AIST’s ABCI-Q — the world’s strongest supercomputer devoted to quantum workloads — integrates an NVIDIA supercomputer with greater than 2,000 NVIDIA H100 GPUs with quantum processors from Fujitsu, QuEra Computing and OptQC.
Elevated availability of those quantum-AI platforms is poised to speed up the breakthroughs researchers could make in quantum computing — together with growing new error correcting codes, integrating quantum processors inside AI supercomputing and simulating low noise designs of quantum {hardware}.
Study extra about NVIDIA’s work advancing accelerated quantum supercomputing at NVIDIA GTC Taipei, operating Could 21-22.
