ResearchIT runs a number of high performance servers which are targeted for solving specific types of problems, and which are managed a bit differently than a traditional supercomputer, and would be considered something between High Performance Computing (HPC) and High Throughput Computing (HTC).
Our servers include names like Speedy, BioCrunch, and BigRAM to give you an idea of their purpose. We allow and encourage interactive use as way to help make the transition from a desktop environment easier. We also allow very long running sessions, sometimes for months, to ensure jobs using less mature code (without checkpoints) can be completed.
The hardware of the ResearchIT servers spans from servers with very fast clock-speed for single or few threaded jobs, to servers with hundreds of cores for applications that need shared memory, but can be well parallelized across many threads. More recently, we have added servers with the latest Nvidia GPUs for applications like molecular dynamics and deep learning. You can see a full list of our hardware here: https://rit.las.iastate.edu/hardware
For research disciplines that do not have decades of history doing traditional distributed parallel computations, the ResearchIT servers are often a better way to get started than jumping right into an HPC cluster.
Our servers are funded as collaborative purchases between researchers, LAS and CALS. If you're interested in using the ResearchIT servers, we'd like to meet with you and explain how things work, please email us at firstname.lastname@example.org. Even if you don't have funds available right now to help contribute to the pool of resources, we will find somewhere for you to get your project done.