Why do you care:
GPUs (graphics processing units) are not required to run basic machine learning (ML) models, but they are often the first choice for training advanced models such as Catboost, Xgboost, neural networks, and deep learning models. Utilizing GPUs on PCs/work laptops brings many advantages such as parallel processing, speed, ability to train with more iterations, ability to train with complex models, and competitive advantage. The example below is from the official Catboost documentation on how to use GPUs to improve training time.
You'll be happy to use the GPU on your laptop and speed up your training. Yes, that's correct. However, you cannot directly take advantage of the GPU until it is enabled. Don't worry, that's exactly what you'll learn from this blog. For example purposes, I will show how to enable the GPU on a Windows OS computer.
let's start.
First, you need to open the device manager on your computer, check your device adapters, as shown in the image below:
You may notice some Nvidia XXX (for example → Nvidia RTX A2000 Laptop GPU). Yes, that's the GPU available on your machine. To take advantage of the RTX A2000 GPU, you need to have the correct dependencies installed on your computer.
list of dependencies/drivers?
NVIDIA Driver
Framework - TensorFlow/PyTorch
Microsoft Visual Studio (MSVC)<