Submer and Intel Partner on AI and HPC Infrastructure Across Middle East, Africa and Turkey
Submer and Intel have announced a collaboration covering artificial intelligence (AI) infrastructure, high-performance computing (HPC) and next-generation data centre deployments across the Middle East, Africa and Turkey.
The agreement brings together Intel’s compute platforms, validated reference architectures and technology roadmap with Submer’s engineering, infrastructure design and deployment capabilities. The companies will work with organisations across the region on requirements ranging from initial infrastructure planning through to deployment and integration.
The collaboration comes as growing demand for AI and HPC capacity is placing greater pressure on data centre infrastructure. Higher compute densities are creating new requirements around power availability, thermal management and the ability of existing facilities to accommodate increasingly demanding workloads.
A key area of the collaboration will be the development of reference architectures for AI, accelerated computing and high-density deployments, including liquid-cooled configurations. The architectures will combine Intel platform technologies with Submer’s infrastructure design and integration expertise, with complete configurations undergoing testing and validation.
Submer will contribute analysis covering areas such as thermal performance, power consumption and infrastructure efficiency.
“Customers in this region are not short of ambition on AI and HPC. What they need is a route from design intent to operating infrastructure that holds together technically,” said Khalid Aljamed, General Manager, Middle East, Africa and Turkey, Submer. “Working alongside Intel lets us pair proven compute platforms with the engineering, integration and delivery capability we have built regionally, so that customers are dealing with one coherent architecture rather than a set of disconnected decisions.”
Addressing high-density deployments
The companies plan to work across sovereign, enterprise, hyperscale, telecom and public-sector environments, using their respective customer and ecosystem relationships.
Submer will also support deployment feasibility assessments in target markets, covering both new data centre projects and upgrades to existing facilities. Retrofitting existing infrastructure is becoming an increasingly relevant consideration as operators look to accommodate higher-density computing without necessarily replacing entire facilities.
The collaboration will also examine infrastructure requirements for AI workloads as organisations move from experimentation towards larger-scale production deployments.
“Organizations across the region are rapidly transitioning from experimental AI pilots to production-scale deployments faster than traditional data centers were built to handle. Today, the strategic focus is fundamentally shifting toward real-world application use cases and continuous inferencing—workloads that demand unprecedented platform agility, thermal efficiency, and compute density,” said Ahmad Alabduljabbar, General Manager, Saudi Arabia, Intel.
“Bringing Intel’s validated reference architectures together with Submer’s advanced immersion cooling and deployment expertise delivers a practical, sustainable pathway to high-density AI and HPC environments. Building infrastructure that flexes to meet both immediate enterprise demands and future technological shifts is critical to realizing the Kingdom’s Vision 2030 ambitions for digital leadership and AI sovereignty.”
Reference architectures and proof-of-concepts
Initial work will focus on reference architecture development and proof-of-concept engagements with customers in selected markets.
Submer will provide infrastructure access, integration and deployment support, while Intel will contribute reference architectures and technical guidance. The companies also plan to conduct joint workshops, executive briefings and technical sessions for customers and ecosystem participants across the region.
The collaboration reflects a broader shift in regional data centre infrastructure, as operators balance the requirements of increasingly compute-intensive AI workloads with constraints around power, cooling, facility capacity and deployment timelines.
