32GB is the sweet spot for Docker.

It helps, but you’re unlikely to notice the difference between 32GB and 64GB for typical local dev.

More RAM is never bad, but Docker containers aren’t the memory hogs people assume. A typical stack – say, a web server, a database, a cache, and a queue – might use 4-8GB total. Even if you run ten containers, you’re usually under 16GB. The bottleneck for Docker on a dev machine is almost never RAM; it’s CPU and disk I/O.

Where 64GB shines is if you’re running memory-hungry services like Elasticsearch, a couple of heavy databases, or multiple full-blown IDE instances alongside the containers. Or if you’re doing something weird like spinning up 50 containers for integration testing. For most web dev, Rails/Node/Python backends and Postgres/Redis don’t demand that much.

If you’re buying a new machine and the price difference is small, sure, future-proof. But don’t spend an extra $400 just for Docker containers. Spend it on a faster SSD or more cores instead.