Saudi-based startup Think has announced it has raised over $8 million in pre-seed funding, marking the largest AI infrastructure and deeptech pre-seed round in the MENA region to date. Co-led by RAED Ventures and Wa'ed Ventures, with participation from Dhahran Techno Valley's Venture Capital arm and strategic angels, the capital is earmarked for team expansion, manufacturing scale-up, and international growth. The company aims to accelerate deployments across Saudi Arabia and the GCC while expanding its global presence. By integrating high-density, liquid-cooled multi-GPU compute nodes with proprietary bare-metal orchestration software, Think seeks to address the rising costs and complexities associated with modern AI infrastructure deployment for enterprises and government organizations.
Think's Integrated Hardware and ILM Software Platform
Think is developing a unified hardware and software infrastructure designed to solve challenges in AI adoption by reducing costs and increasing efficiency. The platform combines proprietary AI Node hardware with ILM, a software orchestration layer. This integration is designed to maximize GPU utilization and lower token costs. In production benchmark testing, the platform achieved sustained GPU utilization exceeding 90%, significantly higher than the industry average of 30–50%. Furthermore, the platform demonstrated a per-million-token cost that is nearly 10x lower than the average cost of using frontier models from Google, OpenAI, and Anthropic.
Crucially, the technology utilizes existing, widely available GPUs and does not require proprietary or specialist inference hardware. The company has indicated that the platform will soon support mixed-vendor and specialist inferencing silicon working in tandem for both inferencing and training. Founded by Ahmed AlSharif, formerly of Meta and Sony PlayStation Europe, and Ammar Enaya, a veteran from Cisco and HPE Aruba, the company is positioning itself as an alternative to the current industry trend of increasing model and data center scale. The company is already engaged in multiple proofs of concept and production deployments across Saudi Arabia, participating in the Kingdom's AI ecosystem alongside initiatives such as HUMAIN.
Scaling Sovereign AI Across Saudi Arabia and the GCC
The company is positioning its technology as the "engine room of the AI era," providing an integrated infrastructure layer for secure, efficient, and sovereign AI deployments. This focus on sovereignty is a direct response to growing enterprise and government demand for infrastructure that provides total control and ownership, avoiding the dependence and security concerns associated with hyperscale cloud providers. Think plans to expand its platform across the GCC over the next 18 months while accelerating the development of ILM as a standalone software platform.
The deployment model is versatile, allowing the integrated hardware, software, and cooling technologies to be utilized across data centers, offices, laboratories, and edge environments. This flexibility is intended to support organizations ranging from startups to large enterprises looking to improve the economics and reduce the environmental impact of their AI operations. As Saudi Arabia pursues ambitions to become a global leader in artificial intelligence, Think is building the foundational infrastructure required to make AI practical, affordable, and sovereign for the region.
Key Takeaways
- Think raised over $8 million in a pre-seed round, the largest for AI infrastructure and deeptech in the MENA region.
- The platform achieved over 90% GPU utilization in benchmark testing, compared to the industry average of 30–50%.
- The ILM orchestration layer can reduce per-million-token costs to nearly 10x lower than frontier models from Google, OpenAI, and Anthropic.
TechInsyte's Take
In our view, Think’s approach signals a strategic shift in the AI landscape from a focus on model scale to a focus on infrastructure efficiency and sovereignty. By achieving 90% GPU utilization using standard hardware, the company is targeting the massive inefficiency currently plaguing enterprise AI deployments. This is particularly relevant for organizations in the GCC that require strict data sovereignty and want to avoid the spiraling costs and vendor lock-in associated with hyperscale cloud providers. The ability to deliver a 10x reduction in token costs while maintaining high utilization suggests that the next phase of AI maturity will be defined by how effectively companies can optimize their existing compute assets. As Think expands ILM into a standalone software layer, it is positioned to move from a hardware-centric provider to a critical orchestration component in the global AI stack.
Source: PRNewswire
