CUDA By the Numbers: 270+ Apps and Counting

by George Millington

CUDA, the parallel programming model that unlocks the power of GPU acceleration, is growing fast.

There are 275 CUDA-based applications tuned to run on GPU accelerators, compared with 90 just three years ago. Most of these applications are household names for researchers and engineers, used every day to accelerate scientific discoveries and engineering results.

Number of CUDA ApplicationsIt’s a list that keeps growing. Some numbers:

  • 738: University CUDA programming courses worldwide
  • 50,000: CUDA-based academic papers published
  • 2,200,000: Downloads of the CUDA toolkit to date
  • 506,000,000: CUDA-capable GPUs shipped worldwide

There’s a CUDA App for That

These numbers underscore how far CUDA has come in a short period of time.

But really, it’s all about the apps. A few examples:

Science: Researchers at University of Illinois Urbana-Champaign used a GPU-accelerated version of NAMD to make a major breakthrough in how the HIV virus is transmitted to healthy cells. They used NAMD, a molecular dynamics application for studying biological matter on the atomic level, for this ground-breaking work. Call out box:

If you would like to try apps like NAMD on the latest, high-performance GPUs, you can test drive GPUs for free.

Entertainment: If you watched the Sochi Olympics on your smartphone or tablet, you’ll probably experience Elemental Technology’s CUDA-powered streaming application. With NVIDIA GPUs, the Elemental Live service can encode 3x more video streams using one-third less energy, helping the company to deliver high-performance feeds to millions simultaneously around the globe.

Engineering: The world’s top computational fluid dynamics software, ANSYS Fluent, is to design more aerodynamic vehicles. With GPU acceleration, engineers can simulate a vehicle’s aerodynamic properties twice as fast, helping them build more fuel-efficient cars and planes.

To find the ideal GPU-supported application for your project, check out the catalog of the latest CUDA applications.