Coastal communities across the United States face a 26% chance of flooding within any 30-year period, a risk that is rising due to sea-level rise driven by climate change. At the University of California, Santa Cruz, researchers are using advanced computing to model and visualize these risks in real time, helping decision-makers plan for a more resilient future. Michael Beck, professor and director of the UC Santa Cruz Center for Coastal Climate Resilience, leads efforts to map how natural defenses like coral reefs and mangroves can reduce flood impacts. His team creates detailed, GPU-accelerated visualizations that illustrate flood scenarios under extreme weather events, enabling governments, NGOs, and financial institutions to assess risks and evaluate nature-based solutions. “We view visualizations as fundamental to motivating action,” Beck said. “These are complex problems involving people and expensive fixes, so we must clearly show how solutions work before investing in them.” To speed up simulations, the team leverages NVIDIA CUDA-X software—including cuPyNumeric and nvfortran—running on NVIDIA RTX GPUs provided through the NVIDIA Academic Grant Program. This acceleration has transformed their workflow. What once took six hours on CPUs now takes just 40 minutes on a single NVIDIA RTX 6000 Ada Generation GPU. With a cluster of four GPUs, they can run four simulations in parallel, cutting processing time by 3 to 4 times, and in some cases even more. David Gutiérrez, senior coastal modeler at the center, noted the impact: “This speedup allowed us to run sensitivity analyses upfront, ensuring we weren’t making incorrect assumptions or setting flawed parameters. It gave us the confidence to scale our work.” The team is now expanding its mission globally, aiming to model flood risks across all small-island developing states—ranging from Tonga to Trinidad and Tobago—before the COP30 climate conference in November. Their high-resolution models use tools like SFINCS to simulate storm events and feed results into Unreal Engine 5, producing interactive, data-driven visualizations. One key application is in coral reef protection. Beck’s work with The Nature Conservancy helped secure a parametric insurance policy for the Mesoamerican Barrier Reef off Mexico’s Yucatán Peninsula. The policy triggers payouts when wind speeds exceed 100 knots during a hurricane, funding reef restoration. This model is now being adopted in the Caribbean and Hawaii, ensuring rapid recovery after storms. “After a hurricane, coral heads break, and waves damage the reef,” Beck explained. “We need to move them to nurseries, reattach them, and clear debris—fishing nets, refrigerators, and other flotsam—brought in by floodwaters.” The team is also advancing the USGS’s Coastal Storm Modeling System (CoSMoS) through a project called CoSMoS ADAPT. This initiative enhances the system’s ability to evaluate the cost-effectiveness of adaptation strategies—from seawalls to restored dunes and nearshore reefs—across California’s coast. “CoSMoS predicts current and future risks,” Beck said. “We’re now adding the solutions that reduce those risks, and doing so at multiple scales. This requires heavy computation and rapid processing—exactly what NVIDIA GPUs deliver.” The NVIDIA Academic Grant Program is currently accepting proposals through December 31 for research in generative AI, AI alignment and inference, and robotics and edge AI. Full-time faculty at accredited institutions are eligible to apply.