Global race to build AI chips, memory, quantum and supporting infrastructure

The global race to develop advanced AI hardware is accelerating at an unprecedented pace, driven by massive capital inflows, strategic geopolitical initiatives, and technological breakthroughs. This surge is shaping the landscape of AI chips, memory, quantum computing, and supporting infrastructure, with a focus on securing supply chain resilience, technological sovereignty, and regional dominance.

Funding and Partnerships Fueling AI Hardware Innovation

A significant driver of this supercycle is the influx of capital and strategic partnerships targeting startups and established players in the AI hardware ecosystem:

• AI chip startups are attracting substantial investments. Notably, MatX, founded by former Google TPU engineers, secured $500 million in Series B funding to develop energy-efficient AI training chips aimed at performance-per-watt optimization. This reflects a broader trend where startups focus on sustainable, high-performance hardware to meet rising AI demands.
• Photonics and data transfer technologies are also gaining attention. Ayar Labs, supported by MediaTek’s $90 million investment, is advancing silicon photonics to enable faster, more energy-efficient data movement within data centers—a critical factor for scaling AI infrastructure.
• Large funding rounds for AI cloud providers like Together AI, which plans to raise around $1 billion, indicate growing demand for massive AI infrastructure deployment, including rental of Nvidia chips and cloud services.
• Governments are actively investing and partnering with domestic firms to build regional manufacturing capabilities. Japan’s Rapidus received an additional ¥267.6 billion (~$2 billion), with the Japanese government acquiring up to 11.5%, aiming to bolster domestic memory fabrication and achieve technological sovereignty. Similarly, South Korea’s SK Hynix is expanding its AI-specific memory chip capacity and supporting domestic innovation by directly purchasing AI startups and fostering ecosystem growth.

Large-Scale Infrastructure and Data Center Investments

Parallel to startup funding, major technology and energy corporations are investing heavily in large-scale AI infrastructure:

• Data center expansion remains a priority for global tech giants. These investments aim to accommodate the explosive growth in AI model training and inference, which requires vast computational resources.
• Energy-efficient hardware is central to these efforts. Startups developing green AI chips—designed to reduce power consumption and carbon footprint—have collectively raised around $500 million. Innovations in silicon photonics further enhance data transfer speeds while minimizing energy use, crucial for sustainable growth.
• Photonics technologies, such as those developed by Ayar Labs, are instrumental in scaling these data centers by enabling faster data movement at lower energy costs, thereby supporting the increasing computational demands of modern AI workloads.

Geopolitical and Resource Security Dynamics

The expansion of AI hardware infrastructure is deeply intertwined with resource security and geopolitical considerations:

• Export controls, such as restrictions on Nvidia’s H200 AI processor, are delaying deployment and limiting access in Chinese markets. China is responding by accelerating domestic memory and chip manufacturing, investing heavily in fabrication plants in Japan, the Netherlands, and other strategic regions to pursue technological sovereignty.
• Critical mineral resources, including cobalt and rare earth elements from Africa’s Democratic Republic of Congo and South Africa, are vital for batteries, magnets, and semiconductors. The scramble for these resources is intensified by regional conflicts, export restrictions, and the need for diversified sourcing.
• Energy security concerns loom large, especially with tensions in the Middle East and Iran’s energy redirection plans. Potential oil price surges above $120 per barrel could hinder the energy-intensive infrastructure projects necessary for AI hardware deployment, increasing operational costs and complicating supply chains.

Implications for the Industry

The confluence of these factors presents both opportunities and risks:

• Supply chain vulnerabilities remain a key concern. Disruptions in chip manufacturing, mineral supplies, and logistics could slow down the hardware supercycle unless mitigated through regional self-sufficiency and diversified supply chains.
• Valuations of AI hardware startups are under scrutiny amid market recalibrations, yet continued large-scale investments by governments and corporations underscore a long-term confidence in the sector’s growth.
• Policy measures are evolving to support domestic manufacturing, resource security, and ecosystem innovation. Countries like Japan, South Korea, and Saudi Arabia are deploying strategic funds and policy reforms to bolster their leadership.

Looking Ahead

From 2024 through 2026, the industry is poised for a technological renaissance. Breakthroughs in photonic interconnects, energy-efficient chips, and regional manufacturing hubs will be pivotal. Success will hinge on stakeholders’ ability to navigate geopolitical risks, manage resource constraints, and scale resilient, sustainable supply chains.

This era marks a strategic pivot toward green chips and self-sufficiency, aiming to ensure long-term sustainability and independence. The race to build AI hardware is not only about raw computing power but also about establishing geopolitical and technological dominance—a contest that will largely shape the global AI landscape for years to come.