WEBVTT

00:00:00 --> 00:03:32
Evolution of general purpose GPU computing and Python

00:03:32 --> 00:05:50
Understanding GPU architecture and massive parallel execution

00:05:50 --> 00:08:18
Exploring the CUDA ecosystem and levels of abstraction

00:08:18 --> 00:11:14
Leveraging domain-specific frameworks and RAPIDS for data science

00:11:14 --> 00:12:39
Replacing NumPy with cuPy for straightforward GPU acceleration

00:12:39 --> 00:14:30
Accelerating math operations and kernel fusion with nvmath-python

00:14:30 --> 00:17:37
Real-world example of supernova identification using Python libraries

00:17:37 --> 00:19:10
Upcoming tools for array-based programming and core compute libraries

00:19:10 --> 00:21:30
Scaling performance across multiple GPUs using specialized frameworks

00:21:30 --> 00:23:04
Profiling and debugging GPU code with Nsight developer tools

00:23:04 --> 99:59:59
Accessing educational resources and the accelerated compute hub