NVIDIA has announced its collaboration with Google Quantum AI to accelerate the design of its next-generation quantum devices through advanced simulations based on the NVIDIA CUDA-Q™ platform. This advancement promises to address key challenges in quantum hardware design and bring the industry closer to developing commercially viable quantum computers.
Supercomputing to Overcome Quantum Hardware Limitations
Google Quantum AI is utilizing the NVIDIA Eos supercomputer along with NVIDIA’s hybrid quantum-classical computing platform to simulate the physics of its quantum processors. This approach aims to tackle one of the biggest challenges facing current quantum hardware: “noise.” This phenomenon limits the number of quantum operations that can be performed before calculations lose accuracy.
“Developing commercially useful quantum computers is only possible if we can scale quantum hardware while keeping noise under control,” explained Guifre Vidal, research scientist at Google Quantum AI. “With NVIDIA’s accelerated computing, we are exploring the implications of noise in increasingly larger quantum chip designs.”
Faster and More Affordable Simulations
Simulating the effects of noise in quantum hardware designs requires extremely complex dynamic calculations that model how qubits interact with their environment. Traditionally, these calculations were prohibitively expensive in computational terms. However, thanks to CUDA-Q and the NVIDIA H100 Tensor Core GPUs, Google has conducted one of the largest and fastest dynamic simulations of quantum devices, employing 1,024 GPUs in the NVIDIA Eos supercomputer. This approach has significantly reduced simulation costs and times.
“The power of supercomputing with AI will be crucial for the success of quantum computing,” said Tim Costa, director of Quantum and HPC at NVIDIA. “Google’s use of the CUDA-Q platform demonstrates the central role of GPU-accelerated simulations in advancing quantum computing to solve real-world problems.”
Quantum Simulations at Unprecedented Scale
With the help of CUDA-Q, Google has achieved realistic and comprehensive simulations of quantum devices containing up to 40 qubits, the largest number reached so far for this type of simulation. These techniques allow for noisy simulations that previously took a week to complete in just minutes, greatly speeding up the development of quantum hardware.
Public Availability for the Quantum Industry
The software used for these advanced dynamic simulations will be publicly available on the CUDA-Q platform. This will enable quantum hardware engineers worldwide to rapidly scale their designs and contribute to the advancement of the field.
The collaboration between NVIDIA and Google Quantum AI underscores the potential of accelerated computing to overcome critical barriers in quantum computing, bringing the industry closer to solving complex problems in areas such as chemistry, cryptography, and artificial intelligence.