From the course: AWS Certified Machine Learning - Specialty (MLS-C01) Cert Prep: 3 Modeling
Unlock this course with a free trial
Join today to access over 23,100 courses taught by industry experts.
Compute choice - Amazon Web Services (AWS) Tutorial
From the course: AWS Certified Machine Learning - Specialty (MLS-C01) Cert Prep: 3 Modeling
Compute choice
- [Instructor] Here, let's talk about how to select the GPU versus GPU for the training job that you're running, or for the inference job. First, let's take a look at a CPU here. It is a cheap and single purpose-built chip. And so what this means is that it's great for doing things like opening up spreadsheets, but it's less great for doing parallel processing because it's not designed to be parallel. On the other hand, a GPU has high density, so it's designed to be parallel, but it has a higher cost and the higher cost allows you to do things in parallel. So there's a trade-off here between a CPU and a GPU and it really depends on what kind of problem you're trying to solve, how fast you want to solve it. And really, this is one of the things to consider, is if you can use a CPU, maybe time isn't something that you care about, it can be done, let's say in batch, and it's not urgent, a CPU could be a good choice. If you…