ORCD Newsletter: May 2026

Above the Fold

Cloud AI Seed Funds

For the past 18 months, ORCD has run a series of three AI seed lab funding rounds. Each round garnered 30-40 proposals. We were able to select 8 each round, based on technical readiness, experience with GPUs, and impact on the overall ORCD environment. We feel like the first two seed rounds, which are now complete, were very successful in highlighting important MIT research, helping our PIs finish important projects, and generating examples and feedback that have guided expanding our services. ORCD currently has raised funds to continue AI seed funds for three more rounds. We are actively seeking funding for follow on activities.

At the same time, we would like to expand our offering of access to hardware for AI-related work by creating a cloud AI seed fund project over the next couple of years. Instead of guaranteed access to ORCD GPUs, the cloud seed fund project would work with a cloud provider to provide research funding support and access to their AI hardware.

There are a few good reasons for this: we'd like to develop a cloud partner overall for ORCD to enable seamless access alongside Engaging. We want to continue to ensure we can enable first-class research at MIT, including through access to capabilities and capacity only available through cloud providers. Through discussions with a couple of providers, we see where we think they can potentially add value. We want to really see who it fits at MIT, and how well.

We at ORCD would love to hear from our community, your thoughts and ideas on this: both how you view access to the cloud and which providers you use now or would like to learn about using. Practically, to support each annual cycle, ORCD has raised $1.5M-$2M/year for the 16 project researchers and compute costs. This is approximately $100-$125K per project that supports researchers and compute and data costs. We think this has been very valuable for ORCD and for researchers. We're going to try to put together a pilot of maybe two rounds of four seed funds on the cloud in the coming few months, so hearing from you now can inform what we're doing and how we can fund it.

To get in touch, email Peter Fisher (fisherp@mit.edu) or Chris Hill (cnh@mit.edu).

What We’re Reading

MGHPCC Power Shutdown, June 15-18

There will be a full power shutdown at the MGHPCC, where ORCD systems are housed, starting Monday, June 15. We will begin to bring our systems down on Monday morning, June 15, at 5am and plan to have them back up Thursday evening, June 18. As components are brought up, they will be made available.

After the shutdown, which is one day longer than most years, the MGHPCC will have 3MW more capacity to support computing - enough to support an additional ~$75M-$100M of computing hardware.

Leading up to the maintenance, you may need to include a time limit for batch jobs to ensure they complete before the start of the reservation for the maintenance at 5am Monday morning. The flag is --time. Please check your system’s documentation for more information.

We will have our usual Office Hours on Tuesday at 10-11am in 46-4199.

Following the MGHPCC power shutdown, we will continue to have regularly scheduled maintenance on the third Tuesday of each month. The next scheduled maintenance is Tuesday, July 21.

If you have any questions please send an email to orcd-help@mit.edu.

Upcoming Workshops and Training

RAG/LLM Workshop

Retrieval-augmented generation (RAG) is a framework for enhancing the knowledge capabilities of a pre-trained large language model (LLM) by providing it with a set of documents to use as a ground source of truth. RAG allows people to utilize the abstract reasoning and summarizing capabilities of LLMs to gain insights on the information provided in a given set of documents. In this workshop, we will learn how to run RAG using the GPUs on the Engaging cluster, as well as tailor the pipeline to work with any set of documents.

This workshop will run on Wednesday, June 24 from 2-4PM.

See the classes page on the ORCD website for more details and signup information.

Introduction to Parallel Programming

In this class, concepts of parallel computing will be introduced. Attendees will learn not only the basics of high-performance computing (HPC) clusters and GPU accelerators but also programming skills with OpenMP, MPI, CUDA, Pytorch, and Deepspeed.

The topics on the two days are independent. Attendees can choose which session(s) they would like to attend.

  • Day One (Wednesday, July 22, 10AM - 4PM):
    • Parallel Programming with OpenMP
    • Distributed Computing with MPI
  • Day Two (Thursday, July 23, 10AM - 4PM):
    • GPU Programming with CUDA
    • Distributed Deep Learning

See the classes page on the ORCD website for more details and signup information.

Introduction to Engaging

The MIT Office of Research Computing's Engaging Cluster is available to the MIT community for running computational workloads that don't run well on your own computer. This hands-on tutorial walks you through the basics of using Engaging for your research.

This course is being run regularly this Spring; the same material will be covered in each instance of the class.



  • Thursday, May 28: 1-3PM
  • Thursday, June 11: 1-3PM
  • Tuesday, June 23: 2-4PM



See the classes page on the ORCD website for more details and signup information.