Nvidia is investing in a quantum computer research center. There, the company will conduct research with partners into useful systems, unsurprisingly coupled with GPUs.
Nvidia is investing an undisclosed sum in a quantum computer research center. The Nvidia Accelerated Quantum Research Center (NVAQC) will be located near Boston. Nvidia will collaborate with partners such as Quantinuum, Quantum Machines, and QuEra Computing.
The purpose of the research center is to further develop both quantum architecture and the associated algorithms. Nvidia hopes to contribute to the development of useful devices.
Changed Course
The announcement at GTC is surprising. Earlier this year, Nvidia CEO Jensen Huang spoke negatively about quantum computers. He called useful applications a future dream that would take twenty years to materialize. At the GTC conference, the CEO takes the stage with partners from the quantum world to explain why that claim was wrong.
It seems that Nvidia has discovered what Nvidia can mean for quantum. It’s not surprising that Huang isn’t enthusiastic about supercomputers based on quantum chips instead of Nvidia GPUs. At GTC, the company now unveils a vision in which GPUs do play a role.
Specifically, the intention is to connect quantum systems to Nvidia GB200 NVL72 Grace Blackwell Superchips via their own Nvidia DGX Quantum architecture. In other words, Nvidia wants to steer the development of quantum systems in a direction where GPU accelerators continue to play a prominent role.
CUDA-Q
Moreover, the company is well aware of what it owes its current success to. On one hand, it’s the very powerful hardware, but on the other hand, it’s also the software ecosystem. When you say GPU acceleration, you essentially mean CUDA in practice. Competitors like AMD have a hard time breaking through with their own solutions, even though CUDA dominance is not popular. Former Intel CEO Pat Gelsinger stated at the end of 2023 that the entire sector wants to eliminate CUDA, but despite various initiatives, it’s difficult to achieve.
As long as CUDA remains the standard, Nvidia’s bread is buttered. With NVAQC, the company is now trying to repeat that trick. Central to the development that Nvidia is backing is indeed the CUDA-Q platform. This software platform is intended to support the development of useful quantum computers.
Thus, Jensen Huang is already gathering quantum partners around him and is starting the gradual development of CUDA-Q. The gamble has a lot of potential: if Nvidia can make its mark with NVAQC and CUDA-Q, the company promises to become intertwined with the development of quantum. With the rise of generative AI, it has become clear how important it is to be the first to offer an extensive software ecosystem to developers.