HUMAIN is stacking up more partners for its Saudi AI infrastructure push, announcing separate deals with AMD, Cisco and Together AI that together underscore its ambitions to become a major hub for AI compute.
AMD, Cisco and HUMAIN said Monday that AMD Instinct MI355X GPU-based infrastructure, connected using Cisco Silicon One-based networking, is now live in Saudi Arabia and serving HUMAIN customers. The deployment uses AMD Instinct GPUs, AMD EPYC CPUs and Cisco’s N9000 Series platform with 800G optics to support GPU-as-a-service for training and inference workloads.
The companies said the next phase of their previously announced joint venture will add up to 250 MW of AI infrastructure beginning in 2027, with capacity expected to start coming online in the second half of that year. That phase is slated to use AMD Instinct MI400 Series GPUs, AMD EPYC CPUs, AMD ROCm software and Cisco networking. The broader joint venture remains on track to deploy up to 1 GW of AI infrastructure by 2030, according to the companies.
Separately, Together AI and HUMAIN announced a strategic partnership at LEAP to connect HUMAIN’s Saudi compute capacity with global AI-native and enterprise demand. That agreement includes construction of a new 250 MW data center in Saudi Arabia and pairs HUMAIN’s multi-chip infrastructure footprint with Together AI’s cloud platform, inference capabilities and developer ecosystem.
For HUMAIN, the announcements show a two-pronged strategy: build large-scale AI infrastructure in the Kingdom while lining up cloud platforms and customers that can use it. The Together AI partnership is expected to generate more than $5 billion in gross annualized revenue in its first year, the companies said. It also highlights a growing bottleneck in the AI market: access to power, data center capacity and low-latency infrastructure close to major demand centers.
Read the full press releases about Cisco and AMD and Together AI here.