JLL, a renowned global commercial real estate and investment management company, has found that the global neocloud segment will potentially expand by a compound annual growth rate (CAGR) of 82 percent between 2021-2025, as competition for AI capacity intensifies and access to GPU resources surges.
Neoclouds are defined as specialized cloud providers offering flexible and on-demand access to graphics processing units (GPUs) critical for AI, blockchain, gaming and scientific workloads. Neoclouds are also known as GPU-as-a-service (GPUaaS) by merit of their ability to allow customers to select tailored solutions and lower costs than hyperscalers through direct hardware partnerships and focused service offerings. JLL finds that this will be due to customers’ demand for flexibility, scalability and cost advantages for specialized AI infrastructure that some traditional data centers cannot adequately support.
“Demand for AI infrastructure is growing at an exceptional pace, and the global data center market has become capacity constrained. Neoclouds have developed an advantage over traditional cloud providers by moving faster and pricing lower with flexible terms. As AI shows no signs of slowing, its success will rely on accessibility to GPU infrastructure, which neoclouds specifically cater to,” says Andrew Batson, JLL Head of Data Center Research, AMER.
According to JLL, a major advantage for neoclouds is their ability to quickly deploy high-density GPU infrastructure versus multi-year builds common in hyperscale data centers. However, the rise of neocloud infrastructure is not expected by JLL to be a detriment to hyperscalers. Given the specialization in AI-workloads that characterize neoclouds, hyperscalers will be better equipped to provide the diverse range of computing services preferred by many global enterprises.
JLL further finds that from an investment standpoint, the risk profile of neoclouds differs from hyperscalers. Neoclouds are characterised by higher capital requirements and sometimes shorter lease terms, creating more immediate risks. However, risks are balanced with the rental rate premium associated with neoclouds versus traditional data center tenants.
“Funding will be a major factor to translate the potential of neoclouds into a reality capable of handling the AI load. Building GPU infrastructure is capital-heavy, and investors should have a clear vision for delivering a viable business model and support from key clients before undertaking an entry into the neocloud space”, said Muhd Syafiq, Director of Data Center Research, Asia Pacific, JLL.
