Jack Dongarra Early Career Award Winner: Staircases and Trampoline: HPC in the Age of

This invited talk marks the presentation of the 2026 ISC Jack Dongarra Early Career Award, chaired by Michela Taufer of the University of Tennessee Knoxville, to Devesh Tiwari, who then delivers a lecture organized around curiosity, purpose, and cricket. On curiosity, Tiwari surveys his group's quantum-HPC systems research: using classical pre- and post-processing to map programs onto qubits while accounting for error-rate variability, reversing circuits and applying AI models to recover correct outputs, and partitioning large circuits into classically approximated pieces with error guarantees. He extends the heterogeneity argument to quantum hardware itself, describing compiler and mapping work for neutral atom, photonic, and superconducting platforms, and applications including quantum machine learning and solving partial differential equations on real quantum hardware. On purpose, he addresses datacenter sustainability, presenting modeling frameworks for carbon, water scarcity, and PFAS impacts, and techniques such as mixing hardware generations, on-site versus off-site water allocation, job deferral, and incentive schemes that reshape user behavior, citing collaborations with RIKEN and Sandia. The cricket theme frames reflections on AI's impact on scientific careers: AI enables steep jumps to high expertise but introduces an AI tax of fluctuating performance, which human mentorship, discipline, and motivation must counteract. He advises early-career researchers to choose problems along axes of cognitive burden and intellectual creativity, and closes by contrasting the staircase model of science with a more joyful, collaborative trampoline model.