About this Webinar
How Shell Runs HPC Across a Hybrid Cloud.
High-performance computing sits on Shell's critical path to revenue. Subsurface imaging makes up 85% of that demand, AI and generative AI another 10%, and workloads like fluid dynamics and reservoir simulation the remaining 5%.
In this conversation, Shell's HPC leaders explain how they've moved past raw benchmarks to focus on price-performance and, more and more, availability. They describe the shift from "lift and shift" data center moves to a hybrid data landscape built with Qumulo and cloud partners, where users don't have to think about where their data lives.
Topics include handling GPU scarcity with a mix of on-prem capacity, cloud capacity, and AWS capacity blocks for bursts, and the current challenge of running projects across regions without high data costs. The discussion closes on Shell's North Star: 10x faster wall-clock time for workflows, with capacity that grows and shrinks as the business needs it while keeping fixed costs low.