From the course: NVIDIA Certified Associate AI Infrastructure and Operations (NCA-AIIO) Cert Prep

Unlock this course with a free trial

Join today to access over 25,300 courses taught by industry experts.

Layer 4: Core libraries

Layer 4: Core libraries - NVIDIA Tutorial

From the course: NVIDIA Certified Associate AI Infrastructure and Operations (NCA-AIIO) Cert Prep

Start my 1-month free trial Buy for my team

Layer 4: Core libraries

“

So, we talked about the operating system driver and virtualization, let us now focus on other component which is layer 4 core libraries. In core libraries, we will talk about two things, how we do programming for GPU and how GPUs would communicate using or what libraries are available for GPUs to communicate that is the topic of discussion. So, we are talking about layer 4 which is about core libraries. I am trying to re-emphasize again, what I have listed here are the most common topics you should be aware for associate level knowledge and for this particular certification but there can be more libraries, more type of drivers and virtualization software may be available depending on your use cases. So let's focus first on the CUDA library, probably you have heard this name multiple time from me but let's now try to explain this particular concept which basically has revolutionized how GPUs are used. So CUDA stands for compute unified device architecture. It is a unified device…

Contents