Overview

From the ACCESS website:

ACCESS is a program established and funded by the National Science Foundation to help researchers and educators, with or without supporting grants, to utilize the nation’s advanced computing systems and services – at no cost.

Whether you’re looking for advanced computational resources – and outstanding cyberinfrastructure – to take your research to the next level, to explore a career in advanced CI or just to experience the amazing scientific discoveries enabled by supercomputers…

BU faculty, students, staff, researchers, and affiliates already have access to BU’s Shared Computing Cluster (SCC) which provides a wide of range of compute resources for many needs. However for those who need specialized hardware/computing environments or for those who need a substantial amount of compute resources (in the tens of millions of CPU hours, terabytes of RAM, or 10-100,000 GPU hours, or more) ACCESS provides a number of options for computing systems.

Sections

 

Getting ACCESS credits

To get running on a system included in ACCESS you will need to submit an application for an allocation of ACCESS credits. There are four different project types each granting a larger amount of credits than the previous one. The comprehensiveness of the application and review process is commensurate with the maximum amount of credits granted by each project type. However it is very easy to start out with ACCESS as the very first project type’s application requires only a paragraph or two of description of what work is planned; these applications are reviewed on a rolling basis, meaning your allocation is typically granted in a matter of days. You can find a summary of each on the ACCESS Project Types page.

To decide how many ACCESS credits might be required for your planned project, you can consult the ACCESS Exchange Calculator to see what resources are available and the conversion rate between credits and actual usage of a particular system. A wide range of systems are available to choose which may appear daunting, however RCS has put together a list of the most common use cases users might have and recommendations for specific systems for those use cases. Refer to the below “Use Cases” section for more detail.

 

Use Cases

The following use cases are the most common situations where you might consider using systems available via ACCESS:

  • Multi-GPU (single-node or distributed) computations
  • MPI/distributed-memory/multi-node CPU computations using thousands of cores
  • Large memory computations needing >1 TB RAM
  • Persistent websites/databases
  • Publicly accessible websites/databases/services
  • Large total amount of GPU hours
  • Large total amount of CPU hours

 

Recommend Systems

 

Last updated: Loading…