How Will AMD, Cisco and HUMAIN Partner on 1GW AI Buildout?

AMD, Cisco and HUMAIN have announced a joint venture to develop large scale AI infrastructure in Saudi Arabia and beyond, beginning with a 100MW deployment scheduled to go live in 2026. 

The companies plan to build up to 1GW of capacity by 2030 by combining HUMAIN’s data centre platform with AMD compute and Cisco networking and critical infrastructure.

The venture builds on an existing collaboration announced earlier this year and reflects Saudi Arabia’s ambition to expand its position in the global AI economy.

HUMAIN, owned by the Public Investment Fund, aims to deliver full stack AI services supported by modern data centre environments optimised for high density compute.

Foundations of the joint venture

AMD, Cisco and HUMAIN will each serve as founding investors and exclusive technology partners for the new venture. HUMAIN will provide the data centre locations and supporting infrastructure, while AMD and Cisco will supply the compute and networking systems that underpin the buildout.

The first phase centres on a 100MW deployment using HUMAIN’s recently developed data centre capacity, AMD Instinct MI450 Series GPUs and Cisco’s critical infrastructure platforms. These components will form the basis of a standardised environment designed for scalable AI workloads.

Dr Lisa Su, Chair and CEO of AMD, says: "Delivering high-performance global AI infrastructure at scale requires strong partnerships. 

“Together with HUMAIN and Cisco, we're combining leadership compute and networking technologies to expand the capacity and global competitiveness of the Kingdom's AI ecosystem. As part of this deepening collaboration, we're also establishing an AMD Center of Excellence in Saudi Arabia to further accelerate local integration and innovation."

The agreement outlines plans to expand capacity beyond the initial sites as demand for large scale GPU resources increases. By securing aligned technology partners, HUMAIN aims to deliver consistent architectures across future phases of development.

Linking data centre growth with national AI plans

The project aligns with Saudi Arabia’s wider plans to build domestic capability for AI, including local compute access, specialist talent and sovereign data capacity. HUMAIN’s targeted buildout is intended to support organisations working on AI applications across sectors such as industry, energy, healthcare and public services.

Cisco’s role focuses on the networking and critical infrastructure layers needed to support high-performance GPU clusters.

Chuck Robbins, Chair and CEO of Cisco, says: "Our expanded partnership with HUMAIN, together with AMD, will mark a pivotal moment in accelerating Saudi Arabia's AI infrastructure. 

“By providing secure, scalable, critical infrastructure for HUMAIN's up to 1GW buildout, Cisco will help the Kingdom turn its vision for a digitally advanced, AI-powered economy into a reality."

The development also responds to gaps identified in Cisco’s AI Readiness Index, which found that while most Saudi organisations intend to deploy AI agents, only a small proportion have adequate GPU capacity. Increasing domestic availability of compute resources is seen as a key step towards meeting this demand.

Building data centre capacity for high-density AI

HUMAIN’s data centres are being developed to support high-density GPU environments, with power and cooling configurations tailored to next-generation compute clusters. The partnership with AMD and Cisco allows these designs to be tightly coupled with the hardware stack they will support.

Tareq Amin, CEO of HUMAIN, says: "Our partnership with AMD and Cisco brings world-class technology and expertise to our mission. 

“Together, we are pushing the frontier of AI compute, bringing up to 1GW of high-performance, cost-efficient infrastructure online in the Kingdom of Saudi Arabia to empower innovators both in the Kingdom and around the world."

The venture’s model pairs data centre infrastructure with defined technology partners rather than using a multi-vendor approach for each site. This is expected to streamline deployment and ensure consistency across multiple facilities as the buildout progresses.

Planned rollout and next steps

Operations for the joint venture are expected to begin in 2026, with the first 100MW phase forming the foundation for further expansion. 

The companies have outlined an ambition to extend capacity to multiple gigawatts over time, although no additional project locations or timelines have been disclosed.

The collaboration formalises a long-term framework for delivering AI-focused data centre capacity, beginning with the initial deployment and expanding as additional sites and power allocations come online.