Nvidia announced Rubin HPC at SC26, a GPU designed to speed up simulations and data processing for science, engineering and analytics. The company says it is intended for workflows where AI agents help researchers test candidate materials, molecules and designs through simulation.
Rubin HPC is part of the Nvidia Vera Rubin platform. It has more than 50,000 CUDA cores on a single GPU, according to Nvidia. The company also says it includes a sparse data acceleration capability to keep those cores supplied with data and uses seventh-generation Nvidia NVLink technology for GPU-to-GPU connections.
Many scientific workloads repeatedly retrieve scattered values from large datasets before performing calculations. Nvidia says Rubin HPC is designed to scale data movement alongside computing capacity, so its cores spend less time waiting for data.
The GPU targets climate and genomic simulations
The company says the GPU targets operations used in simulation and analytics, including sparse matrix calculations, stencil computations, sorting and reductions. Such work appears in applications ranging from climate models and engineering simulations to genomic analysis and financial risk.

Nvidia says its CUDA-X libraries provide optimized algorithms that agents can use with Rubin HPC for simulation and analytics. The company expects faster simulations to give scientists and engineers feedback sooner as they assess possible designs or experiments.
Nvidia expects Rubin HPC to be available in 2028
Nvidia expects Rubin HPC to be available in 2028.



