NASA has published a conference presentation describing the computing systems behind Johnson Space Center’s Flight Sciences Lab in Houston. The document, listed with a November 15, 2026 publication date in the NASA Technical Reports Server, reports two computing clusters spanning nearly 700 servers, more than 10 petabytes of parallel storage, and over 1,000 registered users.
The lab supports analysis across NASA’s human spaceflight programs in more than 20 engineering disciplines. Its work includes calculating spacecraft trajectories, examining aerodynamic forces, modeling heat exposure, and supporting rendezvous and docking analysis. The presentation also describes real-time analytical support to flight operations through the Mission Evaluation Room.
Jeff Triplett, the Flight Sciences Lab lead from Intuitive Machines, and Tim Gregoire, deputy lead of operations from Amentum, prepared the presentation for Supercomputing 2026 in Chicago. NASA’s record distinguishes its September 30 acquisition date from its November publication date. CIQ’s conference program listed Triplett’s session for November 17.
Two Types of Engineering Calculation
The lab combines resources from three earlier high-performance computing laboratories. Its presentation identifies two established workload types: highly parallel computational fluid dynamics and large collections of serial Monte Carlo calculations. These approaches use computing resources differently because the calculations have different relationships to one another.
Computational fluid dynamics models the movement of fluids, including air and other gases. A spacecraft analysis can divide the surrounding flow into many small regions and calculate conditions across them. Parallel computing allows multiple processors to work on those regions, exchanging information as the calculation advances.
Monte Carlo analysis repeatedly runs a calculation with changed inputs. In engineering, that approach helps examine how a range of initial conditions or uncertain parameters affects an outcome. Many separate runs can be distributed across available computing resources even when an individual run proceeds serially.
The presentation identifies Europa and Sirius as the lab’s clusters. Its published technical slides describe more than 28,000 CPUs across the server installation. The storage system supports the large input and output files associated with simulations, and a separate archive includes long-term storage and tape.
NASA’s description of JSC aerosciences places the lab within a unified computing environment supporting the center’s engineering work. Aerodynamic and aerothermal analysis examines forces and heating during atmospheric flight, including reentry. Other listed capabilities address low-density gas flows and the interaction of thruster plumes with spacecraft surfaces.
Workstations, Software, and Remote Access
The computing clusters are part of a larger service environment. The presentation lists more than 100 Linux workstations and over 30 services. Engineers can use physical desktop workstations or remote Linux engineering sessions, including sessions that support three-dimensional graphics.
Open OnDemand provides web-based access to computing resources. GitLab supports software development, and automated runners connect development tasks with the computing environment. Container-based workflows package software and its dependencies so a defined computing task can be run within the intended software environment.
The lab also provides a software license server and web-based project documentation. Those services sit alongside the processing and storage resources used for engineering calculations. Research software engineers help adapt workflows, improve code performance, and configure calculations to run at the required scale.
The work differs from the separate scientific computing objectives recently described in NASA Ames documents. New Space Economy’s coverage of space-weather computing goals examined proposed methods for faster solar simulations. The Flight Sciences Lab presentation describes the shared operating environment used for spacecraft engineering and mission analysis.
Maintaining the Linux Environment
The presentation describes the lab’s transition from CentOS 7 to Rocky Linux 8 and its adoption of Ansible for automated operating-system configuration. Ansible allows administrators to apply defined configuration steps across machines. That approach is relevant to an installation with hundreds of servers and a separate workstation fleet.
The slides also discuss Rocky Linux 9 and commercial support from CIQ. They identify support for containers, hardware differences, core services, graphics, and virtualization. The presentation reports accelerated patching and use of the software environment across computing-node types and workstations.
One example concerns degraded graphics performance and computers hanging during sleep and wake operations. According to the presenters, the problem depended on the hardware configuration and was difficult to reproduce. CIQ obtained matching hardware, sent it to Japan to recreate the configuration, reproduced the issue, and supplied a solution.
The lab’s supported engineering disciplines extend to parachutes, structural response, thermal protection, electromagnetics, and injury biomechanics. Its service list connects those analysis activities with flight operations, including computing support supplied through the Mission Evaluation Room during missions.


