Building AI Into Systems That Can't Afford to Be Late - with Drew Barbier of MIPS

For AI systems built into vehicles, robots, and industrial equipment, being smart isn't enough — the difference between success and failure often comes down to whether real-time performance was designed in from the start or added on afterward. In this episode, Drew Barbier, Vice President, IP Business Unit at MIPS, breaks down why starting with the software workload, rather than the hardware, produces measurably better outcomes in real-time compute design. The conversation covers a real-world example of a three-to-five-times performance gain from workload-first design, what changes on a team's