User Guide

Welcome to the DeepSense cloud computing user guide. You don't have to have background of cloud computing, but it is helpful if you spend 5 minutes reading our wiki page "Introduction to Cloud Computing".

Unlike using on-premises systems, you don't have to have a Linux user account to ssh into a Linux boxes unless your workloads have to deploy to a virtual machine. What you will have are Identity and Access Management(IAM) accounts on AWS, GCP, or Azure. DeepSense will setup virtual machines, severless Machine Learning compute platform, data processing pipeline, and unlimited data storage for users. We will provide step by step guidance to help you develop your Machine Learning or data processing projects in cloud. You don't have to stick to any specific cloud. DeepSense will help you find the best solution to make your project development process straightforward in cloud.

The process of developing your projects are:
1. Users and DeepSense will discuss the background of your projects. Users will provide their technical requirements, for example, data type and size, CPU, memory, GPU, and so on.
2. DeepSense will transfer your data to a cloud storage and set up the permission that only the owner of the data can access. Your data is encrypted both in transit and at rest.
3. DeepSense will set up the environment according to your requirement and create an IAM account for you in the cloud.
4. DeepSense will notify you when the set up is finished and provide detailed instructions how to develop your projects. Using cloud's Platform as a Service (PAAS), you only need to provide your code to train your models or process your data.
5. According to the specific environment set up for you, please select corresponding instructions below to start developing your projects. You can also find the instructions in the section Onboarding and using cloud services in the navigation menu on the left side of this page.

Using Cloud Services (SageMaker Studio, ML Workspace, etc.)

For streamlined development, deployment, and management of your machine learning models, we leverage cloud-based services that offer a comprehensive suite of tools and infrastructure.

Please follow the instructions at "Using AWS SageMaker Studio", "Using Azure ML Workspace".

Using HPC Platforms on AWS and Azure

We're moving beyond traditional virtual machines to unlock the full potential of cloud-based High-Performance Computing (HPC). Enables you to working with demanding workloads that require significant computational power, such as simulations, complex modeling, large-scale data analysis, or AI/ML tasks, HPC platforms on AWS and Azure provide the resources you need.

Please follow the instructions at "Using HPC on AWS and Azure".