PERFORMANCE Signal 356 2 feeds carried it
Fujitsu’s Monaka CPU upgrades branch prediction and ALU count for broader workloads
Monaka increases ALU count to six and adopts a three-level TAGE branch predictor, boosting general-purpose performance while keeping 256-bit vector pipes for HPC.
The upgraded branch predictor reduces misprediction penalties, improving performance on irregular code. Six ALUs allow more parallel integer operations, benefiting workloads that are not vector-heavy. Narrower 256-bit vector units lower power and area, making the chip viable for broader markets.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
Monaka uses a three-level TAGE branch predictor to improve prediction accuracy over A64FX’s perceptron-based predictor.
The core provides six ALU ports, implying at least a 6-wide issue width versus A64FX’s 4-wide design.
Vector pipelines are reduced from 512-bit to 256-bit to save power and better match general-purpose software.
THE CLUSTER
↗