In the basement of Connecticut’s Beneski Museum of Natural History, fossilized footprints of a small, chicken-sized dinosaur lie preserved in stone—each mark more than a relic of the past. These ancient tracks are data in motion, holding secrets about how these creatures once moved across soft, prehistoric mud. Unlocking those secrets, however, requires tools far beyond what existed during the Mesozoic Era: high performance computing. Today, advanced simulations allow scientists to reconstruct each step with astonishing precision—down to the individual sand grains that shaped the impression. This isn’t just about dinosaurs. It’s about understanding the fundamental mechanics of locomotion, with implications stretching into robotics, medicine, and the science of movement itself. Professor Peter Falkingham of Liverpool John Moores University leads this effort, combining paleontology, biomechanics, and HPC to recreate the moment a theropod pressed its foot into ancient sediment. With postdoctoral researcher Ben Griffin, Falkingham is pushing the boundaries of what’s possible by simulating millions of particles in real time, capturing how force, pressure, and substrate deformation interact during a single stride. Falkingham’s background is uniquely suited to this work—his training spans both paleontology and computer science. He first used HPC during his PhD to simulate dinosaur tracks using finite element analysis. “I’ve always been fascinated by how computing can help answer biological questions in paleontology,” he says. Griffin, though new to HPC, quickly found himself immersed in the world of high-performance systems. Coming from a background in PowerShell and Windows CMD, the shift to Linux command lines was initially daunting. But with guidance from Falkingham, he learned to navigate the LJMU “Prospero” Cluster, a system launched in 2020 that supports both short test runs and large-scale simulations across different computing nodes. “It’s been surprisingly smooth,” Griffin notes. “Peter walked me through everything, and we set it up together.” He now advises newcomers to seek mentorship and study existing submission scripts. “Don’t try to figure it out alone,” he says with a laugh. “I once pushed a simulation too fast into the substrate and particles exploded everywhere.” This learning curve highlights a key truth: HPC isn’t just for experts. With the right support, researchers from any background can access powerful tools to drive discovery. The real challenge lies in scale. Each simulation models tens or even hundreds of millions of sand grains—each about a millimeter in size—interacting dynamically under the weight of a dinosaur’s foot. On a regular computer, such a task could take months or years. “The physics of each particle interaction—forces, accelerations, collisions—demands massive computational power,” Falkingham explains. Reducing the number of particles might make it run on a laptop, but at the cost of realism. “Once the grains are too large, they exceed the size of the toes,” he warns. “Then the footprint no longer reflects reality.” Time is another constraint. Simulations must capture motion in tiny fractions of a second, requiring continuous, high-resolution calculation. A single run can take days on the Prospero system, and refining results often means running multiple full-scale versions—no quick tests allowed. What began as a quest to understand ancient movement is now sparking collaborations far beyond paleontology. Falkingham’s team is working with roboticists to study how legged robots interact with soft, deformable terrain like sand or mud. Insights from how dinosaur toes spread and adapt could inform better robotic foot design, improving stability and mobility. The research also holds promise for human health. Walking on sand naturally helps people with hip injuries by providing adaptive support. Could footwear be designed to mimic the passive stability seen in dinosaur and bird feet? Falkingham is exploring that possibility. Even sports science is benefiting. The team is now simulating human movement across sand and mud, helping to analyze gait and force distribution. These unexpected applications show how studying the past can illuminate the present. Beyond science, the work has inspired collaborations with computer visualization experts and art students. At Brown University, researchers developed ways to visualize volumetric tracks in 3D. Local art students used virtual reality to explore the data, transforming scientific output into immersive experiences. Each simulation generates hundreds or thousands of text files—each containing millions of lines of particle data. “What do you do with all of this?” Falkingham asks. The answer is still evolving, but it’s clear that science is no longer just about data—it’s about interpretation, communication, and creativity. For Griffin, the journey has been transformative. “It’s amazing to see how this work connects to so many other fields,” he reflects. “Curiosity is what HPC has truly ignited.” As the HPC community gathers at SC25 in St. Louis this November, the story of dinosaur footprints serves as a powerful reminder: breakthroughs often begin with a single step—replayed, reimagined, and redefined through the power of computation.