Rhino.Compute: Cloud VS Physical Server

Hi!
I’ve been using Grasshopper for a while now and I’d like to store my Grasshopper definitions on a server and run them using Rhino.Compute. I’ll soon need to have a talk with my boss to convince him either to spend a lot of money on cloud renting or spend a lot of money on building a physical server at one of our locations :sweat_smile:

My questions are:

  1. Should I rent a cloud or build our own server?
  2. If I rent a cloud, what service would you recommend (there seem to be many users on Azure and AWS, any thoughts on them?)
  3. Are there any pitfalls I should be aware of concerning cloud services?
  4. If a build a server, what specs should I be looking for for my hardware?
  5. Are there any pitfalls I should be aware of concerning physical servers?

The transition to Rhino.Compute will require a lot of time and resources so any advice that can help me avoid a wasted investment would be greatly appreciated.

1 Like

Sorry, a bit late to the party.

I can just tell you my anecdotal experience with our setup:

  1. Depends on your internal IT department. They should be asked first, if your company has standard procedures for this topic in place (Hosting VMs, Cloud). If you dont have an IT department, i would recomend you a Cloud Provider. The bigger ones (AWS, Azure) are a little pricier. Using a smaller one could save you about half of the amount you would use at the big ones.
  2. My experience with Azure Windows VM was quite smooth for deployment
  3. Not yet encountered.
  4. Rhino uses more CPU then GPU. CPU optimzed VM are luckily cheaper. You should start with the cheapest one, and go from there. If its to slow, get a VM with more CPU.
1 Like