In the latest inference processing MLPerf benchmark round, Gaudi came in a fairly close second to H100. Read More in  Forbes – AI