Harvard Medical School
Research Computing



Status


HMS Research Computing Grid




Projects


Orchestra Cluster

The Orchestra cluster is a shared high-performance computing environment serving a large research community with diverse research requirements and workflows. Tens of thousands of jobs run on the cluster every day, and we are constantly improving the job handling software and configuration to balance throughput between our many users.

Storage

The thousands of researchers we support generate a tremendous amount of data. Some of our group resources and time go towards evaluating future storage products for cost, performance, and suitability for our researchers.

We currently have over 3 petabytes of network attached storage, between the various file servers. Our latest addition is a Lustre based scratch filesystem, which has the potential to deliver very high read/write performance.

Cloud Computing

The computational resources needed by our community wax and wane with the academic year and the funding deadlines. We are looking towards the future and evaluating cloud computing as a possible solution to provide surge compute capacity during times of high need.

Classes

Members of our group are also pro-active in the Quad community and teach classes from time to time. This has ranged from introductory UNIX to Perl/BioPerl as well a more specific introduction to our cluster, Orchestra. We have slides from previous classes available.

Consulting

Have a research problem and not sure how to tackle it? Contact us and have one of our research computing consultants chat with you. Consultants have a life science background and can meet with you one-on-one to help you figure out how to use computers effectively in your research or to help you get access to the resources you need.




Support


Questions about the Orchestra cluster? Fill out this form to get an Orchestra account. The Orchestra wiki has lots of information on using the cluster, including this Introduction to using Orchestra.

Or, pick one of the buttons below to request support. You can either: (1) have the Harvard community help answer your questions through OSQA, (2) fill out a web form with detailed information about your issue, (3) send us an e-mail, or (4) just visit us in person.





People


We are a small team dedicated to promoting and maintaining the use of HPC in the research community. Meet some of our members below.


Andy Bergman

Research Systems Administrator

I help support the HMS user community to better use our HPC, web hosting, and software licensing services.

Doug Feldmann

Senior Research Systems Architect

Doug's focus is on the web-hosting service and base infrastructure of the Orchestra cluster.

Peter Doherty

Research Systems Administrator

I look after the Orchestra HPC cluster, with a focus on optimizing the queuing, operating system, storage, and network components.

Amir Karger

Senior Research Computing Consultant

I love talking to scientists about their research and helping catalyze their way over the computational barriers. "perl" is one of my favorite verbs.

Greg Cavanagh

Senior Research Computing Consultant

My focus is to find what is most broken, and to design and implement new solutions that are not broken.

Chris Botka

Director of Research Computing

I strategically visualize things. Ask me about the future of Orchestra. Suggestions welcome.

Nam

Research Computing Consultant

#bigdata #science #hpc #bioinformatics

David Hummel

Senior DevOps Engineer

Previously worked as a molecular biologist, bioinformatician and software developer. Now designs scalable and maintainable research computing infrastructure, both physical and virtual.

Tom Chittenden

Senior Biostatistics & Mathematical Biology Consultant

I develop mathematical and statistical models to further understanding of complex biological systems.

Lingsheng Dong

Senior Research Computing Consultant

Helping biomedical researchers with their computational needs. Any feedback is welcomed.

We are hiring!

You

Associate Director of Research Computing

I would like to apply for this job.




Downloads


We do have some code available through our Github.

There are many research applications installed on Orchestra. If you don't see a program you want, just ask us.

The following licensed desktop research applications are distributed by research computing; all downloads require HMS eCommons ID authentication. Please note that the HMS Affinity ID and Harvard PIN will not work for this.

  • ChemOffice is a suite of chemistry tools available to all Quad-based faculty, staff, and students of HMS/HSDM, and licensed for use on or off the HMS network. [installation notes]
  • Geneious is a bioinformatics software platform that is licensed for use on the Harvard Medical School network. [installation notes]
  • GraphPad Prism combines scientific graphing, comprehensive curve fitting (nonlinear regression), understandable statistics, and data organization. [info]
  • JMP is statistical discovery software that is licensed to all members and affiliates of Harvard Medical School. [download page]
  • JMP Genomics [info]
  • Lasergene is a suite of sequence analysis software by DNAStar for Windows and Mac, with analysis tools including alignments, contig assembly, gene discovery, primer design, restriction mapping, protein structure prediction, and more. Lasergene is available for use only on the HMS network. [download page]
  • Mathematica is licensed to all members and affiliates of Harvard University. Mathematica can also be run on the Orchestra HPC in addition to the desktop versions. [download page]
  • MATLAB is available under a site license to all Quad-based faculty, staff, and students of HMS. Harvard FAS users can obtain MATLAB from FAS directly. Members of HMS affiliates, such as MGH, BIDMC, Children's Hospital, etc., should seek MATLAB licensing through their primary institution. MATLAB can also be run on the Orchestra HPC in addition to the desktop versions. [download page]

See the HMS Software Wiki for links to other Harvard organizations with software licensing resources.



About


Helping beginners and experts get research done on computers.

Computational research is an increasingly important part of any life science research project. Research computing staff help researchers get access to software and hardware resources and provide advice on how to use those resources. Ask us big or small, beginner or expert questions.

We can help you with things like...

  • access an HPCC (high-performance compute cluster) to analyze data,
  • install or use a particular bioinformatics program,
  • make your programs run faster,
  • reformat data so that a particular program will read it,
  • run MPI parallel jobs on a cluster,
  • run Matlab or Mathematica interactively or in batch,
  • download exon sequences for genes, with 100 bp on either side,
  • run your analysis 5000+ times using a simple script, or
  • get terabytes of backed-up storage for next-gen sequence analysis.

Research computing consultants with a life science background can talk with you about your project, help you decide an effective strategy for the computational part of your research, and acting as liaisons to other parts of IT to access the necessary resources. If we can't help with a particular problem, we'll try to find somebody who can.

All services are offered free of charge.


Orchestra

Orchestra is a shared high-performance computing (HPC) environment maintained by our group. It has approximately 500 compute nodes and roughly 3 Petabytes of network attached storage. Although it was originally designed to enable computationally intensive bioinformatics research, it has evolved to support other IT solutions including web hosting. You can read more about the Orchestra hardware here.


History

By recommendation of the CIO, The Research Information Technology Group (RITG) was founded in July 2003 by the Information Technology department to support research computing on the Quad that falls beyond the scope of local system administrators or the existing HMS IT department's desktop support.

In 2011, by recommendation of the Research Computing Governing Council, RITG was expanded with the addition of a Director and Research Computing Consultants. The new organization was named the Research Computing group (RC).

Besides augmenting research IT support on the Quad, RC works with other departmental administrators to pool resources, efforts, and expertise. We aim to foster collaboration within the community of systems administrators at HMS to share knowledge and establish common best practices.




Contact


You are more than welcome to visit us at the following address at any time. However, we may not always be there so it's best to send us an e-mail to confirm first. We do have weekly office hours on Wednesdays from 1-3pm. Stop by to chat about anything, or bring your data and questions.

HMS Research Computing
Gordon Hall
25 Shattuck Street #500
Boston, MA 02115
(617) 432-2000
rchelp@hms.harvard.edu
We also have a fledgling online social presence on Twitter, Facebook, and Github.


Contact
HMS Research Computing
Gordon Hall
25 Shattuck Street #500
Boston, MA 02115
P: (617) 432-2000
E: rchelp@hms.harvard.edu
Hours

24/7/365 :)