Description
This session will answer a simple question for AU attendees: “In a world obsessed with GPUs, where should I actually spend my money across CPU, GPU, and NPU for my design, make, and emerging AI workloads?†For years, workstation buying has been “GPU‑first.†Agentic AI and modern Autodesk workflows are pushing CPU back to center stage as the orchestrator for both traditional and AI workloads. Traditional workloads (Revit model updates, AutoCAD drafting, Autodesk Fusion modeling, simulation, office, and collaboration) are still heavily CPU‑bound. Piling on GPU does not fix a starved CPU. Agentic AI and AI assistants add a new control layer that is CPU‑intensive orchestration: planning, tool calls, APS integration, retrieval, evaluation, and logging. The right answer is a balanced portfolio of CPU, GPU, and NPU mapped to how AU customers actually work, not just to spec sheets.
Key Learnings
- Learn about agentic AI and how it differs from traditional generative AI in the context of Design and Make workflows.
- Learn about the orchestration role of the CPU versus NPUs and GPUs in local, agentic AI systems for Autodesk workloads.
- Learn how to map example agentic workflows such as design review, design‑to‑make, and project knowledge assistants.
- Learn when to deploy local, Intel-powered agentic AI systems for secure, low‑latency Autodesk workflows versus the cloud.