AMD Helion
AMD's open rack-scale blueprint for agentic AI infrastructure, with CPU-only configurations up to 32,768 EPYC cores and an integrated model-call token router.
Overview
Introduced by AMD on October 12, 2026, Helion is an open blueprint for agentic AI infrastructure based on an Open Compute Project spec and built on AMD compute. Unlike racks designed to host the largest models, Helion is positioned as capacity for executing agents: it lets customers tailor CPU core density, performance, and power to agentic workloads, with an integrated token router that directs agents' model calls across frontier and open-source models. Ecosystem partners are building systems to the spec for planned availability in November 2026.
This is a sourced directory entry. Features and capabilities are reported in the linked sources; this entry does not document independent Aymorphic testing.
Key features
- CPU-only configuration with up to 32,768 cores from sixth-generation AMD EPYC 9006 Series processors
- Enterprise configuration mixes EPYC CPU blades with up to 16 AMD Instinct MI350P GPUs
- Integrated token router directs agent model calls across frontier and open-source models
- Open Compute Project-based open hardware/software specification, not a proprietary single-vendor rack
Use cases
- Hyperscale deployment of agent execution capacity separate from model-hosting infrastructure
- Enterprise agentic AI infrastructure combining CPU agent execution with GPU inference
- Cost/power-tuned infrastructure for running many concurrent AI agents
Pricing
Not a packaged product with list pricing; it is an open rack specification. Ecosystem partner systems planned for availability in November 2026.
Pricing reflects the source record and may change. Confirm current terms with the provider.
Dates and sources
- Released
- 12 Oct 2026
- Record updated
- 13 Oct 2026
- Record last checked
- 13 Oct 2026