From the original article: "Most architecture research papers are based on simulations running small, easily portable benchmarks that project potential perform

From the original article: "Most architecture research papers are based on simulations running small, easily portable benchmarks that project potential performance if ever implemented. This article is not one of them but rather a retrospective evaluation of machines running real, large production workloads in datacenters since 2015, some used routinely by more than one billion people." It's probably also the first time that Google mentions explicitly that the TPU is hosted by an Intel system, which obviously they manage for you. The phrasing around "18-core, dual-socket Haswell processor" is ambiguous and might not be the exact spec for a TPU host: it's not clear if the "platform that is also the host server" refers to just Intel or that specific configuration. Such ambiguity is usually deliberate, of course...