That’s an interesting question. It is a continuation of the work on the Cell in some sense, as the Cell was made up of a six core SoC (plus 2 other cores, one locked, one for the OS), and the initial batch was actually manufactured in 90nm SOI which was a process that Lisa Su contributed to. The Cell also prioritizes throughput rather than single-core speed, a philosophical approach similar to Zen1. It was way ahead of its time, that’s for sure, and was widely used for scientific workloads, but didn’t see much traction beyond that, and it was too difficult to program for (programmers even today typically favor writing code sequentially rather than threading).
But Zen 2 is quite a bit different. It’s x86 (the Cell was RISC), and it’s a lot more flexible in terms of programming. But it does come with many cores as the Cell did, so that part is indeed similar. I wrote and article on Speed vs Throughput in the articles section, might be worth a read 🙂