nvcc does: NVIDIA's CUDA compiler that produces PTX and SASS

NVCC compiles CUDA code into PTX and SASS

NVCC is a proprietary compiler developed by Nvidia specifically for CUDA programming. It translates CUDA code into PTX (Parallel Thread Execution) and SASS (Parallel Thread Execution Shader Assembly) formats, enabling execution on Nvidia GPUs.

Example

A developer writes a CUDA kernel in C++ and uses NVCC to compile it into PTX and SASS, which can then be executed on an Nvidia GPU.

Remember this

Understanding NVCC's role in compiling CUDA code is essential for developers working with Nvidia GPUs to leverage parallel computing capabilities.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

SASS is: the actual machine code that runs on NVIDIA GPU hardware

SASS: compiled machine code executing on NVIDIA GPU hardware

Parallel Thread Execution

PTX is an intermediate GPU instruction set used in Nvidia's CUDA

CUDA

CUDA enables parallel computation on GPUs

a Triton kernel is

Can your phone run faster with a different brain?

TensorRT does: NVIDIA's inference optimizer that quantizes and fuses operations

TensorRT optimizes deep learning inference by quantizing and fusing operations for NVIDIA GPUs

Arm architecture family

ARM processors are the most widely used family of instruction set architectures

Swipe through 100 ML concepts daily

Open Pocket Polymath