AMD Helios
Provider 
Founded 1969
Sells To Consumers, Enterprise
Pricing Model Software, Transactional
Ownership Public

AMD Helios is a rack-scale AI compute system architecture for large GPU accelerator clusters, standardized through OCP.
AMD Helios is a rack-scale reference architecture that groups dozens of AMD Instinct accelerators into a single scale-up domain, designed for large AI training and inference workloads. It has been adopted as an Open Compute Project standard.
The platform integrates custom racks with backplanes, power delivery, and cooling for hyperscale and AI infrastructure deployments. Competing modular platforms argue cabled scale-up offers greater flexibility than integrated rack designs.
| Attribute | AMD Helios | Nonstop AI | Nvidia NVL, NVLink | |
|---|---|---|---|---|
| Provider | ||||
| Founded | 1969 | — | 1993 | |
| Sells To | Consumers, Enterprise | Enterprise | Consumers, Enterprise | |
| Pricing Model | Software, Transactional | Pre-Revenue | Licensing, Product Sales, Subscription | |
| Ownership | Public | Private | Public |
Provider 
Founded 1969
Sells To Consumers, Enterprise
Pricing Model Software, Transactional
Ownership Public
Provider 
Founded —
Sells To Enterprise
Pricing Model Pre-Revenue
Ownership Private
Provider 
Founded 1993
Sells To Consumers, Enterprise
Pricing Model Licensing, Product Sales, Subscription
Ownership Public


AuraLinks is an open-standards network fabric for scale-up and scale-out interconnect in AI data centers.


SkyHammer is a rack-scale, low-latency scale-up fabric designed from the ASIC up for synchronized AI workloads.


Passage is Lightmatter's photonic interconnect platform that uses 3D vertical integration to scale bandwidth density for AI training systems.
AuraLinks is an open-standards network fabric for scale-up and scale-out interconnect in AI data centers.

SkyHammer is a rack-scale, low-latency scale-up fabric designed from the ASIC up for synchronized AI workloads.

Passage is Lightmatter's photonic interconnect platform that uses 3D vertical integration to scale bandwidth density for AI training systems.
