Keynote Speech: Learning Velocity — Accelerating the Future of Semiconductor Innovation
Artificial intelligence is reshaping the semiconductor industry—not only by driving unprecedented demand for compute, but by changing what it takes to deliver next-generation chip performance. For decades, manufacturing progress was defined largely by scaling: shrinking design nodes to increase transistor density. Today, AI chips require system-level advances across all device types, including GPU, CPU, AI accelerators, HBM, DDR, eSSD, SRAM, scale-up and scale-out networks, co-packaged optics, SiC/GaN for PSUs, and more. As complexity rises and technology development cycles accelerate, the industry’s most critical constraint is shifting: success will depend less on capital alone and more on the speed at which the ecosystem can learn.
Learning velocity is a defining capability for the AI era of semiconductor manufacturing, requiring faster innovation, richer information and deeper collaboration. As AI devices push performance and power requirements higher, no single technology or manufacturing step can determine success on its own. Innovation must occur across the broader semiconductor supply chain—from wafers, reticles and materials to logic and memory devices, interposers, IC substrates, advanced packaging and printed circuit boards. The next breakthroughs will come from accelerating learning across these interconnected segments.
Information in semiconductor manufacturing is essential to increasing learning velocity. KLA’s inspection, metrology and data analytics systems uniquely span the semiconductor ecosystem, providing comprehensive, connected data that helps manufacturers identify variations, accelerate yield learning and create manufacturing knowledge at scale. Process control provides the infrastructure that transforms data into actionable intelligence. In this environment, process control is not simply a production requirement; it is a strategic differentiator.
Finally, AI chip performance is increasingly the sum of many parts. Because yield and reliability are shaped across the ecosystem, learning velocity requires deeper collaboration among semiconductor manufacturers and equipment suppliers.
Drawing on KLA’s position across the semiconductor ecosystem, this executive summit presentation will show how accelerating innovation, information and collaboration can help the industry meet the demands of AI. In a market defined by complexity, success will belong to the companies that learn fastest—and turn learning into the engine of semiconductor progress.
