NVIDIA is introducing CUDA Rust for native GPU programming.
CUDA
8 tracked signals on CUDA.
CUDA Toolkit 13.4 introduces support for Windows on Arm.
CUDA Python 1.0 launches with stable APIs, giving Python developers direct full GPU platform access.
NVIDIA CUDA 13.3 introduces tile-based C++ kernel programming with automatic low-level GPU management
“NVIDIA CUDA Tile programming in C++, enables high-level, tile-based kernel development that automatically manages complex low-level GPU details for optimal performance and portability.”
NVIDIA frames its AI infrastructure as 'fungible' — one platform for all accelerated workloads via CUDA
“it maximizes throughput per megawatt to maximize revenue and deliver lowest token cost to maximize your profit margin”
NVIDIA CUDA is essential for GPU-accelerated computing in AI.
NVIDIA nvmath-python bridges Python workflows with CUDA-X high-performance math libraries
NVIDIA CUDA Tile enables high-performance GPU kernel development within existing C++ codebases