Fujitsu has announced the next chip after the A64FX, an ARM processor (opens in a new tab) which is used in one of the most powerful supercomputers in the world, Fugaku.
The new chip from the Japanese tech giant was announced in a presentation within Fujitsu ActivateNow: the pinnacle of technology (opens in a new tab)held at the Computer History Museum in Mountain View, California. Fujitsu CTO Vivek Mahajan said it will be a “next-gen ARM-based processor for DC” expected to be released in 2028.
With the working title “MONAKA”, said Fujitsu Register (opens in a new tab) that he focused on creating a chip with high efficiency and better energy efficiency, saying that he wanted to contribute to “the realization of a carbon-neutral and sustainable society.”
Energy saving
The new processor is said to be capable of boosting HPC workloads as well as providing optimal performance for AI (opens in a new tab) and data analytics applications, all while delivering “overwhelming energy efficiency” compared to rivals. Fujitsu claims it will have 1.7 times the application performance with twice the performance per watt.
MONAKA is part of a program run by the research agency of the New Energy and Industrial Technology Development Organization (NEDO) in Japan. Its goal is to increase energy savings by 40% in national data centers by 2030.
Fujitsu also contributes to this with other solutions beyond MONAKA, such as low-power accelerators, smart network cards (NICs) using photonics and disaggregation technology.
It is speculated that the new chip will therefore be more similar to those used in servers that support cloud storage (opens in a new tab)cloud hosting (opens in a new tab) and collocation providers (opens in a new tab). However, these chips are less concerned with saving power per se than achieving full utilization.
“The next-generation DC processor (MONAKA) we are developing will have a wider range of features and prove to be more energy-efficient,” a Fujitsu spokesperson told The Register. “The range of potential applications is wider than that of the A64FX, which has special features (e.g. interconnects) unique to Fugaku.”
While details are lacking at the moment, it seems likely that the MONAKA will have the same standout features as the A64FX, such as a 28Gbps Tofu-D interconnect, high-speed HBM2 stack memory, and 512-bit scalable vector extensions.