In a recent report by TD Cowen, it was found that approximately 2.1GW of data center leases were signed in the past 90 days as a result of increasing AI requirements. This surge in leasing activity highlights the growing demand for data center capacity and has significant implications for the industry.
Current state of the US data center market
TD Cowen’s report estimates that the third-party US data center market currently stands at around 10GW. This gives us perspective on the magnitude of the leasing activity observed in the last 90 days, indicating its significance. The increasing adoption of AI technologies is driving the need for more data center capacity, necessitating these lease agreements.
Growing demand for data center capacity
The report further reveals that there are multiple requirements for over 500MW from hyperscalers, as well as at least one high-demand requirement from the US government. This highlights the strong demand in the market for data center resources. Additionally, hyperscalers have started pre-leasing capacity much earlier than before, with a lead time of 24-36 months, compared to the 12-18 month pre-leasing window seen in the previous year. This trend signifies increasing competition and a sense of urgency among hyperscalers to secure future compute access within data centers.
Increasing Scarcity and Tightening of the Leasing Market
The leasing market for data centers was already relatively tight in 2022, and it has tightened drastically in 2023. This scarcity can be attributed to several factors. Firstly, hyperscalers are keen on securing their access to future compute, recognizing the significance of data center capacity in the AI-driven era. Additionally, AI workloads are less sensitive to latency and can be located virtually anywhere in the country, further intensifying the demand for data center resources.
Effects on supply chains and lead times
The surge in data center leases has had a significant impact on supply chains. With increased demand for equipment orders, supply chains are expected to suffer. The report suggests that once these leases translate into actual equipment orders, lead times are likely to further increase. This can be challenging for data center providers, as well as for companies relying on timely access to data center resources for their AI-driven operations.
Implications for Equinix
Equinix, a leading data center provider, is expected to benefit in the medium term from the expanding demand for inference workloads. The build-out of AI technology and the increasing reliance on data centers for AI-driven operations present favorable opportunities for Equinix. With its robust infrastructure and network connectivity, Equinix can cater to the growing needs of customers requiring AI-powered data processing and analysis.
The recent surge in data center leases driven by AI requirements underscores the growing demand for data center capacity. With multiple high-demand requirements from hyperscalers and the US government, the leasing market has tightened significantly. This surge not only impacts supply chains but also suggests a scarcity of data center resources. As the industry adapts to these changing dynamics, companies like Equinix are well-positioned to capitalize on the growing demand for AI-driven workloads. However, it is crucial to address the potential challenges associated with longer lead times and supply chain disruptions to ensure the smooth operation of the data center ecosystem.