Beyond CUDA - GPU Accelerated Computing on Cross-Vendor Graphics Cards with Vulkan Kompute - AMD, Qualcomm, NVIDIA & Friends

Explore GPU accelerated computing across multiple vendor graphics cards using Vulkan Kompute in this 29-minute conference talk. Dive into the cross-vendor GPU compute ecosystem, learning how to leverage general-purpose GPU computing capabilities on AMD, Qualcomm, NVIDIA, and other GPUs. Discover how to write a simple GPU-accelerated machine learning algorithm from scratch that can run on virtually any GPU. Get an overview of projects enabling cross-vendor GPU acceleration and learn to harness your GPU's full power using the Kompute framework with just a few lines of Python code. Gain insights into optimizing through lower-level C++ interfaces and understand the components of Vulkan Kompute. Follow along as the speaker demonstrates setting up Kompute logic, creating tensors, initializing parameters, and executing the main sequence for a linear regression model. Conclude with a high-level roadmap for future developments in cross-vendor GPU computing.

Intro
Hello, my name is Alejandro
High level Objectives
Why Parallel Processing?
Parallel Processing: Options
Vulkan C++ SDK Disadvantages.
Enter Kompute The General Purpose Vulkan Compute Framework.
Vulkan Kompute: Components
The Hello World of ML
ML Example Intuition
Kompute Logic to Set Up
LR Shader Logic
Kompute Logic: Create Tensors
Kompute Logic: Init Tensors
Kompute Logic: Main Sequence
Kompute Logic: "Learn" LR Params
Kompute Logic: Print LR Params
High level Roadmap