Localized climate modeling, new materials for renewable energy, patient-specific medicine – all rely on the availability of massive compute power fed with large data sets, and increasingly, on the integration of traditional simulation techniques with artificial intelligence to solve previously intractable problems. To accelerate these blended workloads, we need to unify diverse architectures of CPUs and GPUs with open and consistent programming models to enable developers to break-free from siloed ecosystems.
We will provide latest workload results on Intel’s existing product portfolio for HPC and AI and will discuss the silicon, software, and system-level innovations designed to minimize “energy-to-solution” and maximize research throughput for the most demanding workloads. Special guests join in-person to discuss how Intel and the open ecosystem are accelerating deployment of these innovations to achieve a future of AI-accelerated HPC.
