Riyadh, August 31 — AMD, Cisco and Saudi‑based AI specialist HUMAIN announced that an AI infrastructure built around AMD’s Instinct MI355X GPUs and Cisco’s Silicon One networking is now live in Saudi Arabia, delivering GPU‑as‑a‑service to HUMAIN customers in the Kingdom and beyond. The production system combines AMD Instinct MI355X graphics processors with AMD EPYC server CPUs, all tied together by Cisco’s N9000 Series networking platform. The Cisco Silicon One and 800‑gigabit optics create an AI‑optimized fabric designed for scale, low latency, and operational resilience as workloads move between GPUs.
That fabric lets HUMAIN offer services—from model training to inferencing—through a GPU‑as‑a‑service model, with the underlying hardware physically located within Saudi Arabia, giving customers direct, sovereign access to advanced compute.
Roadmap for Scaling
Looking ahead, the three partners plan to broaden the platform with a next‑generation suite of AMD Instinct MI400 Series GPUs, paired again with AMD EPYC CPUs and AMD’s open‑source ROCm software stack. The expansion will be powered by the same Cisco networking and critical infrastructure. Deployment of this larger‑scale infrastructure is slated to begin in 2027, with capacity expected to start coming online in the second half of that year.
The joint venture’s long‑term ambition is to deploy up to 1 GW of AI infrastructure by 2030, positioning the region as a hub for sovereign, high‑performance AI workloads.
Strategic Significance
“Bringing AMD Instinct systems online in Saudi Arabia is an important milestone in the work with HUMAIN, building an open, high‑performance AI platform serving customers today and scaling significantly in the coming years,” said AMD Chair and CEO Dr. Lisa Su. HUMAIN CEO Tareq Amin added that the production infrastructure demonstrates the company’s ability to build, operate and deliver advanced AI infrastructure at scale, noting the growing demand for sovereign, high‑performance AI capabilities.
“The region is moving from AI investment to delivery,” observed Cisco Chair and CEO Chuck Robbins, highlighting Cisco’s role in constructing the critical networking foundation that turns AI vision into reality. Here’s what it means for governments, enterprises, research institutions and developers: an open platform that blends open models, open software and locally operated infrastructure, offering a ready‑made environment to train sophisticated AI systems without relying on overseas data centers.
What’s Next
All eyes will be on the rollout that begins in 2027, as the joint venture scales the infrastructure and brings additional capacity online. The pace of deployment will indicate how quickly the Kingdom can meet rising demand for AI services, while the open‑software approach promises broader accessibility for innovators worldwide.
