Solve challenges with powerful GPUs. Develop skills in high-performance computing and apply them in many fields.
Suggested by: Coursera (What is Coursera?)
No prior knowledge required
No unnecessary risks
The internship is designed for data scientists and software developers who are interested in creating software that takes advantage of available hardware. Students will be exposed to CUDA and libraries that allow for multiple calculations to be performed simultaneously and quickly.
Learners will do at least 2 projects that will allow them to explore CUDA-based solutions for image/signal processing, as well as a topic of choice that can be related to their current or future professional career.
They will also create short demos of their efforts and share their code.
What you will learn: Students will learn to develop parallel software in Python and C/C++ programming languages. Students will gain a basic level of understanding of GPU hardware and software architectures.
What you will learn: Students will learn to use the CUDA framework to write C/C++ software that runs on Nvidia CPUs and GPUs. Students will transform algorithms and sequential projects into CUDA commands that execute hundreds to thousands of times simultaneously on GPU hardware.
What you will learn: Students will learn to develop software that can be run in computational environments that include multiple CPUs and GPUs. Students will develop software that uses CUDA to create interactive GPU computational code for handling asynchronous data.
What you will learn: How to develop software that performs advanced mathematical operations using libraries like cuFFT and cuBLAS. How to use the Thrust library to perform a variety of data manipulations and data structures that hide memory management. How to develop machine learning software for a variety of purposes using neural networks modeling the cuTensor and cuDNN libraries.
