AMD Expands Saudi AI Infrastructure to 1GW by 2030

AMD Scales Middle East Data Center Presence

Advanced Micro Devices has put its first Instinct MI355X GPU-based systems into production in Saudi Arabia through an expanded partnership with Cisco Systems and HUMAIN, a regional AI infrastructure provider. The live deployment now serves customers both within the Kingdom and internationally, offering GPU-as-a-service for AI training and inference workloads.

The production infrastructure pairs AMD Instinct MI355X GPUs and EPYC CPUs with Cisco Silicon One networking and 800-gigabit optics. But the current deployment represents just the opening phase of a substantially larger buildout.

250MW Expansion Planned for 2027

The three companies are preparing to deploy up to 250 megawatts of additional AI infrastructure beginning in the second half of 2027. This next phase will utilize AMD's upcoming Instinct MI400 Series GPUs alongside EPYC processors and ROCm software, integrated with Cisco networking and supporting infrastructure.

The partnership maintains its target of deploying up to 1 gigawatt of AI infrastructure by 2030, positioning AMD to compete more directly with NVIDIA and Broadcom in large-scale data center deployments. The platform is designed for governments, enterprises, research institutions, and AI developers requiring locally operated infrastructure with control over data residency and model governance.

Why It Matters

Sovereign AI—where nations build domestic computing capacity to maintain control over data and models—represents a strategic growth vector for chip makers as geopolitical concerns drive infrastructure localization. AMD's HUMAIN partnership gives it a foothold in a market where its open-model approach and data residency controls may prove more attractive than alternatives to government buyers. The scale of the commitment also validates AMD's Instinct GPU roadmap at a time when hyperscaler diversification away from single-vendor dependencies is accelerating.

Data Center Momentum Continues

The expansion builds on strong momentum in AMD's Data Center business, which generated $6.7 billion in revenue during the second quarter of 2026—a 107 percent year-over-year increase driven by EPYC processor demand and Instinct GPU adoption. AMD has noted that Instinct sales more than doubled year over year as deployment broadened across AI labs, cloud providers, startups, and national laboratories.

The company expects server revenues to grow more than 70 percent in 2027, with total Data Center segment revenues projected to more than double year over year. Growing deployments of Helios and MI450-series systems are expected to drive significant acceleration in 2027.

The HUMAIN project specifically targets the sovereign AI segment, where AMD reported already-increasing Instinct adoption. As countries invest in domestic AI infrastructure to retain control over computing capacity, AMD's position as an alternative to dominant players could strengthen, particularly in regions prioritizing vendor diversity and open software ecosystems.

These details were first reported by AI Watch.