Of course, you can't compare those directly with epyc/xeon because of the architectural/instruction set differences. EDIT: meaning that a larger sized cache wo

Of course, you can't compare those directly with epyc/xeon because of the architectural/instruction set differences. EDIT: meaning that a larger sized cache wouldn't necessarily deliver better performance (though I think in this case the architectural differences do favor Graviton making better use of cache)