- Intersect360 Research’s analysis delves into the HPC-AI market, revealing a $85.7 billion size in 2023.
- Hyperscale companies’ AI infrastructure spending surged by 170.2% to $48.4 billion.
- On-premises HPC-AI market reached $37.3 billion, growing by 7.1%.
- Hyperscale AI demand witnessed volatility, with steady growth predicted from early 2026.
- On-premises HPC-AI infrastructure anticipates double-digit growth in 2024, followed by a dip in 2025 due to GPU supply adjustments.
Main AI News:
In a recent analysis released by leading analyst firm Intersect360 Research, fresh perspectives emerge on the significant value of the HPC and AI market. Combining these two pivotal technologies into a singular narrative, Intersect360 Research reveals that the HPC-AI market achieved a remarkable size of $85.7 billion in 2023.
The report underscores a staggering 62.4% year-over-year surge in market size. This surge is primarily attributed to substantial investments by hyperscale entities in their AI infrastructure, nearly tripling compared to rates seen in 2022.
Hyperscale enterprises, defined by their focus on internet-driven data access, processing, and distribution, are pivotal players in this narrative. Their service spectrum spans web search, online retail, content hosting, social media, and beyond.
Intersect360 Research, long renowned for its scrutiny of the hyperscale domain, has recently merged AI with HPC and hyperscale into a unified market. According to the report, the integration of AI with HPC infrastructure is pursued by hyperscale giants, traditional HPC clients, and non-HPC enterprise users alike, catering to both training and inference needs. This convergence necessitates a holistic comprehension of the HPC-AI infrastructure industry due to the intertwined requirements of these once-distinct buying segments.
With a consolidated perspective, Intersect360 Research delineates two pivotal segments: Hyperscale entities and on-premises HPC-AI infrastructure.
Hyperscale AI infrastructure investment soared to $48.4 billion, marking a monumental 170.2% surge from the preceding year.
“This unprecedented surge in hyperscale AI adoption resembles a modern-day gold rush,” remarked Intersect360 Research CEO Addison Snell. “It defies conventional market dynamics, reflecting a convergence of diverse stakeholders charging into the AI arena simultaneously.”
However, the report notes that elevated prices prevail due to shortages in critical components, notably GPUs essential for major computational tasks in training and inference.
Conversely, the on-premises HPC-AI infrastructure market expanded to $37.3 billion in 2023, representing a modest 7.1% uptick compared to 2022. Notably, 2023 marked the first year where hyperscale AI infrastructure spending surpassed the entire expenditure on on-premises HPC-AI infrastructure.
The report emphasizes that each segment constitutes only a fraction of a broader landscape, with on-premises computing and hyperscale infrastructure extending beyond AI applications.
On-premises HPC-AI witnessed a marginal uptick to 30.9% of total on-premises expenditure, compared to 28.6% in 2022. In contrast, the AI segment of hyperscale infrastructure more than doubled to 28.8%, from 14.2% in 2022.
Looking ahead, Intersect360 Research anticipates a stabilization in market growth. On-premises HPC-AI infrastructure is projected to grow at a 5.4% CAGR, reaching $48.6 billion by 2028. However, this will be offset by a decline in hyperscale AI spending, anticipated to experience a -4.6% CAGR. By 2028, hyperscale spending is forecasted to reach $38.2 billion, resulting in an overall .3% CAGR for combined HPC-AI infrastructure spending.
Despite the overarching trend towards stabilization, the journey towards 2028 is envisaged to be volatile. On-premises HPC-AI will witness double-digit growth in 2024, driven by sustained demand for AI systems. Yet, a subsequent dip in 2025 is expected as GPU supplies catch up. Steady growth is anticipated to resume in 2026.
Conversely, hyperscale AI demand will exhibit volatility typical of a market segment dominated by a few key players. Intersect360 Research foresees a moderation in demand for AI infrastructure in the near term, with steady growth returning in early 2026.
Conclusion:
Intersect360 Research’s findings underscore the burgeoning importance of AI integration within both hyperscale and on-premises HPC infrastructure. While hyperscale AI investment surges resemble a modern gold rush, the market faces challenges such as GPU shortages. Nonetheless, the anticipated stabilization and subsequent growth trajectories suggest a promising future for the HPC-AI ecosystem, albeit with expected fluctuations in demand and supply dynamics. Adaptation to these evolving trends will be imperative for stakeholders aiming to capitalize on the expanding opportunities within the HPC-AI market.
