Is Moore's Law Really Dead?

Moore's Law is introduced as a comforting or possibly unsettling prediction about computing power doubling roughly every two years, but the video questions whether the traditional framing still holds water. It traces the origin of the law to Gordon Moore and Fairchild Semiconductor, recounting how Moore helped lay out the idea that the number of components on a chip would double on a regular cadence, and how that notion evolved as Intel and other companies built the modern semiconductor industry. The discussion then pivots to the ambiguity in defining transistor size and density, noting that different manufacturers measure differently and that three dimensional designs and new materials complicate a simple count of transistors per area. Interviewees highlight that even with physical limits acknowledged by Moore in 2005, the industry has found other levers beyond simple transistor scaling to drive performance, such as chiplets, new transistor architectures, and the rising role of AI and ML accelerators. The speaker also situates the debate in a broader tech context, mentioning costs, wafer economics, and the inevitability that silicon performance gains may slow, but that innovation in design and architecture could still deliver meaningful advances in areas like automobiles, gaming, and media fidelity. The overall takeaway is that Moore's Law as a blanket predictor may be fading in its original form, while the industry continues to pursue multiple paths to improve computational capability and value for users. The video closes by underscoring that even if doubling transistor counts stalls, hardware and software ecosystems will still push toward increasingly capable and diverse computing experiences.

Topics · science and technology · computer science · semiconductors · technology history