How Does Fractile Build AI Chips in House?

Walter Goodwin says Fractile builds AI chips in house with about 150 people who cover workload understanding, front-end design, physical design, back-end implementation, and advanced packaging. The group stays skinny in each area and runs one agile closed loop instead of handing a finished stage to an outside partner.

The loop starts by running ahead of the chip. The team looks at where a model architecture should change so it lines up with the Fractile bet, and judges the scaling-law direction of the bets being taken. They then advocate for the direction with customers and partners. The same work informs a set of bets for a chip that is not necessarily going to exist in volume for one or two years, and that foresight is a core muscle they have to build.

Older chip projects followed a waterfall and a handoff point. Reach a certain level, pass the work to another partner, and then live at the mercy of how the partner behaves. AI chips have to chase the tail of the workloads faster than earlier chip plays, so the bets have to land correctly and the team has to strike very fast. The mix takes a lot of skill and a lot of luck. Keeping the entire story in house is how Fractile matches the cadence the industry now sets.