Cloud Research Computing
Cloud Research Computing
The storage and computing resources offered by public cloud vendors can provide benefits of scale, cost savings, and security to research groups. Research Computing and Data Services can partner with you to pick the right tools and services and help you smoothly transition to the cloud.
Data Storage
Cloud providers offer a range of services for data storage from long-term archives to high performance file systems for your computing. For example, you may find affordable pricing to keep your idle data for years or create a secure and reliable space for your regulated data. If your application can benefit from cloud storage, we can help you determine if the solution is right for you and establish data pipelines and lifecycles as needed.
Computing
The basic computing element in the cloud is a virtualized computer, commonly referred to as a virtual machine (VM) or a compute instance. The resource abstraction via virtualization combined with varying hardware properties provides many instance types that provide flexibility to meet various needs, serving a broad range of computational workflows with specific requirements. Several examples of how researchers may take advantage of cloud computing are listed below:
Overflow Computing
Some time-sensitive workloads require large computing resources for a short period. If Northwestern’s HCP cluster, Quest, cannot scale to your needs, Research Computing and Data Services can help you to launch additional resources in the cloud.
Secure Computing for Regulated Data
Research Computing and Data Services has a variety of methods to build VM workstations, high performance computing clusters, and other batch computing systems in the cloud to ensure data security for HIPAA or other regulated data. These secure environments are created to provide both compliance and reliable performance with the required software environment.
Powerful Interactive Computing
Some data sets are too large to analyze on a standard laptop or a high-end workstation; some data sets or computing capabilities may be needed only for a short time. In these circumstances, investing in a powerful workstation may not be justified. For instance, you may have a machine learning model that requires a GPU, but only infrequently and for only a few hours at a time. We can help design a solution that allows you to pay-per-use for the resources you need, including high memory, high throughput, and GPU-enabled instances, and provide guidance on how to use it to take advantage of the capabilities of the cloud.
On-demand Workshop or Course Environments
On-demand workshop or course environments in the cloud are sometimes the ideal solution for a training that requires flexibility for applications, number of users, amount of computing power etc. Research Computing and Data Services staff can help to set up an environment that meets your educational needs, create user accounts, and install necessary software. We have a library of on-demand environments that support software such as R and SQL. If the software needed in your course is not already supported by existing environments, we can work with you to design a new one.
Managed Databases
SQL and NoSQL databases are enormously helpful tools for analyzing data, but the process of creating one is tedious and error prone. There are several options for pay-per-use database services across multiple cloud vendors. Please contact Research Computing and Data Services if you are planning to use a managed database and need help to determine the best option.
Northwestern Cloud Account Benefits
Northwestern has signed agreements with the following public cloud providers:
A Northwestern account established through these agreements allows to bill purchase orders directly in NUFinancials (rather than credit cards or other forms of payment) and provides appreciable cost benefits for cloud services. Learn how to set up your Northwestern cloud account.
Get Started
To learn more about cloud options or meet with a Northwestern IT Research Computing and Data Services consultant, please request a consultation.