You want to run some heavy tasks in a cloud using GPUs. What would you do?
- What features are the most important for you in a GPU cloud provider? Is it price, availability, GPU models, or something else?
- How do you choose an instance type to run? Typically there are dozens of different instances in each provider.
- Do you regularly use one or more providers?
When considering platforms for machine learning operations, I lean towards GCP or Azure. They both offer straightforward MLOps solutions, and I’m well-knowledges in their infrastructures.
Questions: 1. Their financial and time efficiency in complex tasks. 2. Each machine type on such platforms comes with a detailed specification sheet. Usually, I do some sort of calculation that allow for the correct machine type selection. It’s important to understand that machine learning often requires substantial VRAM and usually it’s your main spec to look for. For instance, training a 7 billion parameter model would need approximately (7b*4)/2^30 GB of VRAM. A lower-level A100(40GB) GPU might suffice for this, but for larger models, you’d need a higher-end A100(80GB) to accommodate all the data. Consequently you need to be aware of your requirements. 3. Personally, I prefer Google’s Vertex AI, Colab Pro, and local development.
I find azure terrible for ml in general. They basically force you on databricks, azure ml studio just sucks compared to gcp vertex.
We’re now on teradata for mlops, and it’s surprisingly ok. High entry cost but overall a lot cheaper than what we used to have on azure/databricks, faster, and better.
We are forced on azure at work, but I used vertex for hobby projects