Christopher Misra is the Vice Chancellor and Chief Information Officer at the University of Massachusetts Amherst. With more than 25 years of experience at UMass, he leads the university’s IT strategy and operations, focusing on building reliable systems and supporting the campus’s teaching, research, and public service goals. Chris also serves as Chair of the InCommon Steering Committee and is a board member for NEREN and OSHEAN. A UMass Amherst graduate with a degree in civil engineering, he approaches his work with a practical mindset and values contributing where it counts.
Research computing has been with us for decades. What’s changed dramatically is the velocity and the shape and sources of the demand.
Generative AI has broadened expectations around what’s possible with data. It’s no longer just computational physicists asking deep questions. Faculty in fields like journalism, linguistics, and archaeology are increasingly turning to computational methods to advance their work.
Questions like: Can I run a longitudinal analysis on ten years of interviews? Could I model how language shifts during crises? — are now part of the everyday conversation in disciplines that once had limited exposure to research computing.
This growing and diversifying interest, paired with increasing competition for research funding, puts institutions at a clear inflection point.
What we build next — technically, strategically, and collaboratively — will determine who gets to participate in discovery, and who gets left behind.
The Case for Accessible HPC
Accessible research computing means more than logging into a cluster. It means rethinking how we onboard, support, and scale use across a broader range of disciplines and individuals.
Some researchers are entirely comfortable in Linux environments, working from the command line and compiling code on the fly. But for many others, especially in fields like the arts, social sciences, or humanities, those tools are unfamiliar, and sometimes intimidating. If our systems require advanced technical knowledge to even get started, then we’re putting up unnecessary gates to entry.
That’s where interface layers like Open OnDemand and JupyterHub become essential. They provide a more intuitive, web-based experience that mirrors how many researchers already work in other software environments. This doesn’t dumb down the tools; it widens the aperture of who can use them.
We should also be expanding how we define “research computing.” It’s not just for the physical sciences. Whether it’s reconstructing ancient artifacts with 3D rendering or running machine learning models on historic text archives, computing is now embedded across the academy. Our infrastructure must meet that reality head on.
UMass Amherst: Upgrading the On-Ramp to Discovery
At UMass Amherst, we’ve seen this shift firsthand. In response, we recently completed a 400-gigabit campus network expansion, funded by the NSF’s Campus Cyberinfrastructure (CC*) program, which connects us to national and international research backbones. It’s a foundational upgrade, and it couldn’t have come at a better time.
One of the primary beneficiaries of this expansion is the NET2 cluster, used in collaboration with CERN’s ATLAS experiment at the Large Hadron Collider. As ATLAS transitions to its high-luminosity phase in 2026, data demands are expected to increase by a factor of five to ten. That’s not a theoretical need; it’s a concrete requirement to stay relevant in a global scientific collaboration.
But just as critical as the infrastructure are the people who built it. Engineers like Jessa Westclark and Daniel Uribe, both part of our team, are at the heart of this effort. Daniel started with us as a student employee and has since grown into a network leader helping to support multi-terabyte data flows. Jessa, through early mentorship and support from the Women in Network Security (WINS) program, has become a national figure in advanced networking.
These are not side stories — they are central to the future of research computing. We often talk about bandwidth, but the people are what is most vital. And in both cases, intentional investment pays off.
Leadership as Infrastructure: Unity, Harmony, and Strategic Capacity
None of this work happens in isolation. And it certainly doesn’t start with servers.
At UMass Amherst, we launched our research computing platform, Unity, with a relatively small but targeted investment of $150,000 from our central IT budget. That seed funding allowed us to test a hypothesis: if we build a service grounded in usability, shared ownership, and value, will people use it?
They did. Unity has since grown into a system-wide service supporting all UMass campuses, as well as several colleges in the region. What started as an infrastructure project became a platform for collaboration.
Critically, the partnerships that allowed Unity to scale were already in place. Whether through the Massachusetts Green High-Performance Computing Center (MGHPCC), our regional networks (UMassNet and NEREN), or colleagues at the University of Rhode Island, the groundwork had been laid. That trust turned into action.
This is the real role of leadership: identifying shared purpose, aligning around mission, and investing — not just in tools, but in people and relationships that make the tools usable.
The Path Forward: From Project to Ecosystem
So where do we go from here?
First, we need to start thinking beyond projects. Too often, IT investments are framed around discrete deliverables and timelines. That approach doesn’t account for the interconnected reality of modern research. Infrastructure investments should be designed as part of ecosystems — where one upgrade enables another, and where multiple efforts reinforce each other over time.
Second, we need to embrace pre-competitive collaboration. At UMass, we compete with other institutions for students and grants. But we don’t need to compete on whether we both maintain a data center. Shared infrastructure, like MGHPCC, allows us to direct resources toward innovation instead of duplication.
Third, we need to invest in talent development with intention. That might mean sending an engineer to a WINS event instead of a mass vendor training. That might mean promoting from within, even when external hires seem faster. It’s not always the obvious choice, but long-term, it builds a pipeline that’s resilient and deeply aligned with institutional values.
Finally, and most importantly, CIOs and campus leaders must stay close to the work. Talk to researchers. Sit in on workshops. Understand what energizes your teams. In an environment with limited resources, we can’t afford to bet on the wrong priorities. Alignment — between strategy, passion, and practicality — is everything.
Building Inclusive Infrastructure: Discovery Without Bottlenecks
Research computing isn’t just about hardware or bandwidth. It’s about equity, access, and building ecosystems that enable people to do their best work. The future of discovery will be shaped not by who has the most powerful tools, but by who makes them accessible and usable across disciplines, institutions, and roles.
Whether you’re a CIO, IT leader, or other key decision-maker, here’s where to start:
- Design for usability. Lower the barrier to entry with intuitive platforms like Open OnDemand and Jupyter that mirror how researchers already work.
- Invest in talent pipelines. Support early-career professionals through mentorship, training, and leadership opportunities, especially with student tech hires.
- Collaborate pre-competitively. Identify shared infrastructure opportunities with peer institutions to avoid redundant spending and accelerate innovation.
- Think in ecosystems, not projects. Connect infrastructure upgrades to broader strategies including cybersecurity, data governance, teaching, and research support.
- Stay close to the work. Engage directly with researchers. Understand their pain points and ambitions. Let their needs shape your priorities.
Discovery shouldn’t be gated by outdated systems or siloed thinking. Let’s build research infrastructure that doesn’t just compute—but includes, connects, and empowers.