Citigroup: Ultra Micro Computer (SMCI.US) performance forecast shows that both revenue and gross margin are significantly higher than expected, and the sustainability of AI demand will be the focus of the financial report

The Zhitong Finance App learned that the initial quarterly results announced in advance by artificial intelligence (AI) server manufacturer Ultra Micro Computer (SMCI.US) on Tuesday greatly exceeded Wall Street expectations, indicating that the AI infrastructure investment boom is still driving rapid growth in server demand. Boosted by this news, as of press release, US stocks surged more than 15% in overnight trading on Tuesday.

Ultramicrocomputer said in a statement that the revenue forecast for the first quarter of the 2027 fiscal year ending September 30 is between 18.5 billion and 19.5 billion US dollars, which is significantly higher than the analysts' average estimate of 15 billion US dollars; based on the median value of 19 billion US dollars in the guide range, the estimated revenue is about 26.7% higher than market expectations. Meanwhile, the company expects gross margins of between 12% and 12.5% for the quarter.

Although Ultramicrocomputer has yet to release full quarterly financial data, initial revenue performance already reflects strong demand in the AI server market. As an important supplier in the field of AI servers, ultra-microcomputers benefit from global technology companies continuing to expand AI data center construction and strong demand for high-performance computing equipment. The company's servers are widely equipped with Nvidia (NVDA.US) chips, and Nvidia GPUs have become the core hardware for training and running artificial intelligence models in most AI data centers today. As large technology companies and cloud computing service providers increase investment in computing power infrastructure, demand for ultra-microcomputer-related products continues to be boosted.

In response, Citi pointed out in a research report released on Tuesday that ultra-microcomputer's preliminary results for the first quarter of fiscal year 2027 showed that both revenue and gross margin were significantly higher than expected — the estimated revenue was 18.5 billion to 19.5 billion US dollars, far higher than the previous guidance of 14.5 billion to 15.5 billion US dollars, and also higher than FactSet's forecast of 14.9 billion US dollars; the gross margin is expected to be 12.0% to 12.5%, which is also significantly higher than the previous 10.4% to 10.8% guidance.

Citi pointed out that ultra-micro computer management attributed the performance exceeding expectations to stronger demand for AI platforms and a more favorable customer structure and product portfolio. The updated performance outlook means that the company's revenue increased by about 280% year over year, while the implied year-on-year increase under the previous guidance framework was about 195%. Furthermore, implied operating profit is about 50% higher than current market expectations, and earnings per share (EPS) are about 60% to 70% higher than current expectations.

Compared to the update released on July 21, the positive signals conveyed by this performance forecast are even more obvious. The July 21 update shows a combination of lower revenue expectations and higher profit margin expectations, while this update shows that both revenue and gross margin significantly exceeded expectations.

Although the ultra-microcomputer is expected to provide more details during the November 3 earnings call, Citi reminded investors that the company's management had already sent a positive signal at the September CitiTech conference. Specifically, the company emphasizes that it is possible to drive growth and profitability through solutions optimized for specific applications. These solutions include CPU systems that support intelligent AI workloads, DCBBS deployment, enterprise-grade AI infrastructure, and larger factory-level solutions. In these areas, ultra-microcomputers can provide differentiated system-level value, not just standardized server hardware.

Overall, Citi believes this announcement is a positive signal for the broader server original equipment manufacturer (OEM) ecosystem, which includes Dell Technologies (DELL.US) and HPE (HPE.US). The bank further confirmed this view in recent OCP-related exchanges: AI infrastructure is increasingly becoming a rack-level system challenge, not just an opportunity at the server level.

As enterprise customers, sovereign AI customers, and emerging cloud service providers (neocloud) deploy larger AI environments, procurement decisions are increasingly revolving around integrated rack-level architectures. This architecture streamlines deployment processes, increases operational efficiency, reduces time to productivity, and optimizes total cost of ownership (TCO). This trend should benefit OEMs with comprehensive capabilities in system design, power delivery, liquid cooling, network integration, deployment services, and infrastructure orchestration.

Looking ahead to the upcoming financial report, Citi expects ultra-microcomputers to focus on discussing the sustainability of AI demand, especially in the context of investors' concerns about the financing status and deployment schedule of emerging cloud service providers. At the same time, companies also need to respond to questions about the intensity of working capital expenditure and ability to convert cash.

Citi believes that a key focus of investors' discussions remains when free cash flow will begin to improve, and by what extent. Investors are looking for evidence to confirm whether the company's strong revenue growth, backlog of orders, and the development momentum of the AI infrastructure business can be converted into continued positive cash flow.

Citi gave the ultra-micro computer a “neutral” rating in the research report, and the target price was 45 US dollars. The target price is based on the bank's application of an 8-fold price-earnings ratio (P/E) to the 2028 forecast value of ultra-microcomputers, which is lower than the company's median price-earnings ratio of 14 times over the past three years.

Citi's comparable group of companies consists of enterprise hardware companies, including original design manufacturers (ODMs). The bank pointed out that compared with peers, ultra-microcomputers apply lower valuation multiples, mainly because their profits are more volatile, profit margins are lower, and customer concentration risks are greater. Furthermore, it is expected that customer due diligence will become more strict, and supplier compliance restrictions will also become more stringent; these factors may cause some suppliers to restrict the supply of parts and cause competitive losses due to recent lawsuits.