AMD Unveils Helion Open Rack With 32,768 EPYC Cores For Agents
TL;DR
- AMD Helion scales to 32,768 Zen 6 Venice cores in a CPU-only OCP rack, with eight named OEM partners shipping November 2026.
- Two configurations target distinct buyers: a CPU-only hyperscale rack and an all-in-one enterprise rack mixing EPYC nodes with MI350P GPU nodes.
- Three Venice CPU binnings match agent tiers: dense 400W for personal agents, balanced 600W for enterprise, high-frequency 96-core for technical agents.
AMD said it will ship a rack-scale reference design next month that packs up to 32,768 EPYC CPU cores into a single rack. The design, called Helion, is pitched as the unit of agentic-AI buying rather than another chip SKU, and in its announcement AMD frames it as an open blueprint for agent fleets.
"Agentic AI is not a single workload. It spans personal agents, enterprise applications, code execution and complex workflows, each with different compute requirements," said Dan McNamara, SVP and GM of Compute and Enterprise AI.
Helion ships in two shapes. A CPU-only configuration uses 6th Gen AMD EPYC processors and scales to the 32,768-core number for hyperscale customers running agent orchestration. A combined all-in-one version pairs EPYC CPU nodes with AMD Instinct MI350P GPU nodes for enterprises that want one rack to carry both the agent control plane and the model inference. Both variants come air- or liquid-cooled, and AMD has bolted an integrated token router into the stack to route agents' calls across frontier and open-source models.
The partner list is wide. Cisco, Dell, HPE, Lenovo, Supermicro, Hon Hai, QCT and Wiwynn are developing configurations. Supermicro's statement specifies ORv3 compliance; the AMD release does not pin the overall rack to a specific OCP revision.
The configurations are planned for November 2026, and AMD said it will contribute the Helion specification to the Open Compute Project. A reference design donated to OCP is an invitation to the ODMs to build without paying for a proprietary rack-level stack, which is a posture Nvidia's current rack-scale products do not take. Against the 457 AI-infrastructure stories our tracker has logged in the last 90 days, Helion stands out as an open-rack blueprint written for agent orchestration rather than for training throughput.
What others are reporting
-
AMD Newsroom Read →
First-party source names all 8 OEM partners (Cisco, Dell, Hon Hai, HPE, Lenovo, QCT, Supermicro, Wiwynn) and confirms ORv3 OCP compliance of the contributed spec.
AMD Helion brings AMD EPYC performance, open software and flexible rack designs together. — Dan McNamara, SVP and GM, Compute and Enterprise AI
-
Tom's Hardware Read →
Tier-1 hardware press confirms spec; headline updated same day to correct ship date from 'next year' to 'later this year,' with URL slug preserving the original read.
-
StorageReview Read →
Details specific Venice SKUs (EPYC 9996, 9956, 9686F), three CPU node binnings mapped to agent tiers, and the Verano low-power Zen 6 host processor slated for H2 2027.
"Agentic AI is not a single workload," said Dan McNamara, SVP and GM of Compute and Enterprise AI at AMD.
-
Futurum Group Read →
Analyst Brendan Burke frames Helion as the first rack-scale AI product treating GPUs as optional, inverting three years of GPU-centric infrastructure norms across the industry.
-
Wccftech Read →
Adds 102 Tbps rack networking bandwidth and Caliptra RoT hardware security compliance alongside the dual CPU-only/enterprise config split and expanded 1H 2027 availability window.
"Agentic AI is not a single workload. It spans personal agents, enterprise applications, code execution and complex workflows, each with different compute requirements."
Originally reported by stocktitan.net
Read the original article →Original headline: AMD Debuts Helion Rack Reference With 32,768 Zen 6 Venice Cores for Agentic AI, Ships November