A new report from analyst firm Intersect360 Research sheds new light on the value of the HPC and AI market. Taking these two technologies as a single unit, Intersect360 Research found that the HPC-AI market size reaches $85.7 billion in 2023.
The report states that the size of this market is up 62.4% year-over-year. The cause of this increase appears to be large-scale enterprise spending on their AI infrastructure, which has almost tripled from 2022 rates.
The hyperscale market is made up of organizations whose business is focused on accessing, processing and disseminating information via the Internet. Large-scale businesses often provide services such as web search and categorization, online retailing and customer identification, content hosting and distribution, social media, communications, pattern recognition, massively multiplayer online games (MMOG), etc.
Credit: Intersect360 Research
Although Intersect360 Research has been following the hyperscale market for many years, the organization recently decided that AI combines HPC and hyperscale into a single market. According to the paper, AI with HPC infrastructure is sought after by large-scale enterprises, traditional HPC customers, and non-HPC enterprise IT users, for both training and inference. There is a need for a combined understanding of the HPC-AI infrastructure sector due to the overlapping needs of these formerly distinct purchasing segments.
With this combined market view, Intersect360 Research focuses on two main segments: hyperscale enterprises and on-premises HPC-AI infrastructure.
Spending on large-scale AI infrastructure increased to $48.4 billion, a massive 170.2% increase from the previous year.
“This dramatic growth in AI at scale is akin to a gold rush,” said Addison Snell, CEO of Intersect360 Research. “This belies any normal market dynamics. It is the confluence of several groups who are simultaneously launching the race for AI.
The report goes on to mention that prices are currently high here due to the shortage of many important components. Specifically, the GPUs needed to provide the majority of computation for training and inference are increasingly difficult to acquire.
In contrast, the on-premises HPC-AI infrastructure market reached $37.3 billion in 2023, up 7.1% from 2022. The report mentions that 2023 was the first year in which Large-scale enterprise spending on AI infrastructure has exceeded total spending on on-premises infrastructure for HPC-AI.
The report also makes a point of mentioning that each of these segments is only a part of a larger whole and that there is both on-premises IT infrastructure and non-AI hyperscale infrastructure.
Intersect360 Research revealed that on-premises AI HPC increased to 30.9% of all on-premises spending, representing a marginal increase from the 28.6% represented in 2022. On the other hand, the share of AI in hyperscale infrastructure increased to 28.8%. more than double the proportion of 14.2% in 2022.
Looking ahead, Intersect360 Research predicts a slowdown in market growth. The on-premises HPC-AI infrastructure market will grow 5.4% to $48.6 billion by 2028. However, this will be offset by a decrease in large-scale AI spending, which will see a CAGR of -4.6% in the coming years. years. By 2028, large-scale spending will reach $38.2 billion. Overall, this will result in a CAGR of 0.3% for combined HPC-AI infrastructure spending.
Despite the flattening of the overall outlook by 2028, the report predicts a volatile development towards this end. On-premises HPC-AI will see double-digit growth in 2024 as current demand for AI systems remains high. However, this will be followed by a decline in 2025 as the increase in GPU supply catches up. The report predicts that steady growth will resume in 2026.
Alternatively, demand for large-scale AI will experience the kind of volatility that can be expected from a market segment made up of relatively few companies. Intersect360 Research predicts a slowdown in demand for AI infrastructure in the coming years, with a return to steady growth in early 2026.
Of course, this is just a taste of what Intersect360 Research has written on this topic, and those with a deep interest in where the HPC-AI market is heading should be sure to download the full report. Additionally, Intersect360 Research will also present the information from this report during a free webinar on Thursday, May 2 at 12:00 p.m. Eastern/9:00 a.m. Pacific. Snell will be joined by Dan Olds, Research Director of Intersect360 Research, as well as Principal Analyst Steve Conway to discuss the insights contained in the report.
Related
