Newapproach

Welcome to Bridges

Bridges is a uniquely capable resource for empowering new research communities and bringing together HPC and Big Data. Explore the virtual tour showing Bridges’ components, how they’re connected, and how they interact for various workloads. Select any part of the architecture for a summary of what it is and how it contributes. Access to expanded detail on the Bridges system hardware, software and user account allocations. Through the Pittsburgh Research Computing Inititative, faculty and staff at CMU and Pitt who haven’t used Bridges yet can try it free of charge.

Virtual Tour User Guide Research Computing Initiative


Bridges FB1b

The Approach to Bridges

Bridges is designed to support familiar, convenient software and environments for both traditional and non-traditional HPC users. Its richly-connected set of interacting systems offers exceptional flexibility for data analytics, simulation, workflows and gateways, leveraging interactivity, parallel computing, Spark and Hadoop.

 

Getting Started

laptop 2Apply to Bridges
Getting started on Bridges is easy; just choose your on-ramp.

networking 3Account Administration
Check your usage, add a user, request VMs
 and other account specific functions.

hierarchical structureHelp 
Check our FAQ for more information.

 

directionsUsing Bridges
Visit the User Guide for detailed instructions on
 accessing and using Bridges.

 

 


settings 1

Tools & Services

Interactivity: Unlike traditional HPC systems, Bridges enables you to work interactively as you would on your laptop.

Virtual MachinesVirtualization provides a dedicated space for interactivity and allows you to customize your software and environments.

Gateways: Gateways allow you to launch jobs, orchestrate complex workflows and manage data from your browsers. 

 

Databases: Bridges’ dedicated database nodes power persistent relational and NoSQL databases so they are always available to you.

Spark/HadoopUnlike traditional supercomputers, Bridges supports Hadoop and Spark, modern tools that enable processing of large data sets in distributed computing environments.

Web Servers: Web Server Nodes host gateways to community applications and provide access to community datasets.


GPUsGraphical Processing Units accelerate a wide range of application.

Extreme Memory: Large shared memory for memory-intensive applications such as machine learning, graph analytics, genome sequence assembly and in-memory databases.

SoftwareA robust collection of applications and tools allows you to focus on analysis instead of coding.

 

Data CollectionsYour research is supported by rapid access to both public data collections and those you host on Bridges for the use of your project or collaboration.

Networking: PSC’s experts can optimize the transfer of your large data sets to Bridges’ available Petabytes of storage at no charge.

User SupportThrough PSC’s own consulting staff and the XSEDE ECSS program, experts with mastery in a variety of HPC areas stand ready to help you advance your research.


peopleCommunities/Use Cases

The communities represented here are not all inclusive of the communities we expect Bridges to serve. Because Bridges is a uniquely capable resource it will empower new research communities and bring together HPC and Big Data in new ways. How will you use Bridges? Here are a few possibilities: Regular memory nodes; Persistent databases; AI and deep learning software; Extreme shared memory nodes; Hosted data collections; Gateways; Application packages; GPUs; Hadoop and Spark environments for data mining and machine learning tasks.

Digital Humanities | take advantage of these Bridges’ features:

  • Persistent databases of various types: relational databases for structured data; document-oriented databases for text; and graph databases  to represent the relationships between texts or authors or both
  • GPU nodes and Hadoop and Spark environments for data mining and machine learning tasks
  • Contemporary applications and a rich software environment including R and python, for text manipulation, statistical calculations, and the graphical display of results
  • Extreme shared memory nodes to manage big data, useful, if not essential, to the success of many digital humanities projects.

Social Sciences | take advantage of these Bridges’ features:

  • Gateways that allow intuitive access to complex workflows

  • Graph analytics that take advantage of both the large and regular memory nodes; the extreme memory nodes can manage larger and larger amounts of data

  • Easy access to data collections  for analysis

  • Persistent databases enable data discovery

  • Access to contemporary applications like R, Java and Python

AI Machine Learning | take advantage of the following Bridges’ features:

  • Popular applications like Caffe, Theano, Torch, scikit-learn, cuDNN
  • Public data collections

  • Large memory to allow manipulation of large datasets
  • GPU enabled tools for deep learning

  • GPU nodes and Hadoop and Spark environments for data mining and machine learning tasks

Physical Sciences | including astronomy, chemistry, physics, geology and more take advantage of the following Bridges’ features:

  • More than 750 “regular memory” nodes with 128GB RAM each provide the ability to run large numbers of jobs, facilitating optimization and parameter sweeps.

  • Persistent databases of various types: relational databases for structured data; document-oriented databases for text; and graph databases  to represent the relationships in your data

  • Contemporary applications like R, Java and Python and a rich software environment with variety and breadth of packages

  • AI and deep learning software such as Theanos, Caffe

  • Extreme shared memory nodes to manage larger and larger amounts of data

  • Hosted data collections that can be mined by researchers

  • Gateways that allow intuitive access to complex workflows

  • GPUs for accelerating a wide range of applications

 Biology/Genomics | take advantage of the following Bridges’ features:

  • Extreme memory nodes which make managing the TB of data required by genome assembly possible

  • Gateways allowing intuitive access to complex workflows

  • Hosted data collections which provide easy access to public data sets

  • Variety and breadth in application packages to accommodate user needs

Neuroscience | take advantage of the following Bridges’s features:

  • Persistent databases of various types: relational databases for structured data; document-oriented databases for text; and graph databases  to represent the relationships in data

  • Contemporary applications and a rich software environment

  • Extreme shared memory nodes to manage larger and larger amounts of data

  • Hosted data collections that can be mined by researchers

  • Gateways that allow intuitive access to complex workflows

Business/Economics | take advantage of the following Bridges’ features:

  • More than 750 “regular memory” nodes with 128GB RAM each provide the ability to run large numbers of jobs, facilitating optimization and parameter sweeps.

  • Persistent databases of various types: relational databases for structured data; document-oriented databases for text; and graph databases  to represent the relationships in your data

  • Extreme shared memory nodes to manage larger and larger amounts of data

  • Hosted data collections that can be mined by researchers

  • Variety and breadth in application packages including contemporary applications like R, Java and Python

Public Health | take advantage of the following Bridges’ features:

  • More than 750 “regular memory” nodes with 128GB RAM each provide the ability to run large numbers of jobs, facilitating agent-based  and ensemble modeling

  • Persistent databases of various types: relational databases for structured data; document-oriented databases for text; and graph databases  to represent the relationships in your data

  • Contemporary applications and a rich software environment

  • Extreme shared memory nodes to manage larger and larger amounts of data

  • Hosted data collections that can be mined by researchers

  • Gateways that allow intuitive access to complex workflows

Computer Science | take advantage of the following Bridges’ features:

  • More than 750 “regular memory” nodes with 128GB RAM each provide the ability to run large numbers of jobs, facilitating optimization and parameter sweeps.

  • Persistent databases of various types: relational databases for structured data; document-oriented databases for text; and graph databases  to represent the relationships in your data

  • Contemporary applications and a rich software environment 

  • Extreme shared memory nodes to manage larger and larger amounts of data

  • Hadoop and Spark environments for data mining and machine learning tasks