I do a lot of data science tasks in Python and R and 8GB is definitely a show stopper a lot of the time. It makes life easier when I can just write naive code t

I do a lot of data science tasks in Python and R and 8GB is definitely a show stopper a lot of the time. It makes life easier when I can just write naive code that loads all of our training data into memory. I like to work locally when possible but often have to move things to our 512GB servers for speed. For GPU heavy stuff I'm running things on our Nvidia machines only but lots of non-GPU tasks are still memory bound. 16GB or better 32GB would allow me to go to the server probably 25% and 50% fewer times, respectively. Which would be great!