… Nvidia unfolded its datacenter roadmap out to 2027 in June 2024 , when we learned about the “Vera” CV100 Arm server CPUs and the “Rubin” R200 GPU accelerators for the first time. And then Huang folded out another year and showed us the datacenter roadmap out to 2028 at the GTC conference last year. …
… Above and below these columns of V3 cores, there are what look like a different sort of core – perhaps these are “Hermes” Neoverse N3 cores, which have been allocated to do more generic server work where vector math is not as important. See the Neoverse roadmaps for more on this . …
… Another key is that the device has three memory tiers: SRAM inside the chip, HBM on the chip package, and DRAM on the accelerator card and – importantly – not on the host server. …
… Scout is optimized for inference, and it is designed so it can fit inside the compute and memory budget of an Nvidia H100 GPU accelerator. Maverick is designed to fit inside of an eight-way Nvidia HGX H100 GPU system board commonly used in server nodes inside of Meta Platforms and other companies. …