Скачать или смотреть How much faster is AI running on a GPU vs a CPU? Let's find out.

How much faster is an AI model when running on a CPU vs GPU? Why is there such relentless demand for faster and larger graphics cards? In this video, we'll explore how to answer this question on your own local computer using a simple combination of Ollama, Docker, and some basic app code written in a language like C# or Python. We'll see how to run an AI model using Ollama in a Docker container to control which resources the AI model leverages, such as primarily the CPU or the GPU. We'll then see how to measure the AI response time in these different scenarios using a few types of prompts sent and tracked by some simple app code. This should be a fun exercise, so please hit those subscribe and like buttons if you enjoy the content.

00:00 Intro & context
01:21 Prerequisites
02:45 Local AI setup on Docker
07:07 App code tour
09:08 Speed test demo & results
16:52 Closing thoughts

GitHub repo: https://github.com/alex-wolf-ps/ai-sp...

Related AI and coding videos:

Use AI to chat with a SQL Database 4.0 - local AI, MySQL and Postgres, and more!
   • Use AI to chat with a SQL Database 4.0 - l...  

Use the new .NET AI project templates to build AI apps!
   • Use the new .NET AI project templates to b...  

Use vector search in Azure SQL or SQL 2025
   • Use Azure SQL and MSSQL 2025 vector search...  

Build an AI app to search your images
   • Build an AI app to search your images usin...  

Build a fully local AI RAG solution with .NET
   • Build a fully local AI RAG app - as simple...  

The .NET AI Ecosystem as simple as possible
   • The .NET AI Ecosystem explained as simple ...  

.NET 9.0 Blazor Features in 9 Minutes
   • New .NET 9.0 Blazor Features!  

Build AI Apps with Semantic Kernel
   • Use Semantic Kernel to build AI Apps and A...  

Setup Stable Diffusion locally:
   • Stable Diffusion AI - a simple intro. Set ...  

AI Upscale any video on your local machine
   • AI Upscale Any Video on Your Local Machine...