According to CoinMeta, AMD today announced the launch of the Helion intelligence AI rack solution, which adopts an open infrastructure design based on the Open Compute Project (OCP) specifications. The aim is to meet the computational power requirements for large-scale AI intelligence tasks and to support flexible combinations of CPU and GPU. The solution offers two main configurations: one is a pure CPU rack equipped with sixth-generation AMD EPYC processors, supporting up to 32,768 CPU cores, designed for ultra-large-scale deployments; the other is an integrated rack that combines AMD EPYC CPU, AMD Instinct MI350P GPU, and general-purpose computing nodes, suitable for enterprise-level AI applications. AMD states that Helion supports both liquid cooling and air cooling configurations, allowing customers to adjust CPU and GPU resources according to their workload, performance, power consumption, and cost requirements. The solution also integrates open-source software and hardware, as well as Token routers, enabling the allocation of intelligence model calls to both cutting-edge and open-source models. AMD plans to submit the Helion specifications to OCP for further development and adoption by the community, with a expected launch in November 2026.