Will demand for AI servers and the order backlog support revenue growth in FY2027-H1?

Revenue in FY2027 Q1 was ¥502.3 billion, up 20.7% year on year, setting a new quarterly record. Components were the main driver of growth, with revenue of ¥349.3 billion, up 27.5%. Data-center-related revenue also rose to ¥69.7 billion, up 81.1%, making clear that demand for AI servers and related equipment is driving company-wide performance.

Among leading indicators of demand, Q1 orders received reached a record ¥673.9 billion, while the order backlog was ¥617.8 billion and the book-to-bill ratio was 1.34x. Production also increased to ¥519.0 billion from ¥427.0 billion a year earlier, so it is important to determine whether multiple indicators, including orders, production and revenue, are improving simultaneously rather than shipments merely being brought forward. In FY2027-H1, we will check whether orders received exceed the ¥673.9 billion recorded in Q1 alone and whether the order backlog builds from ¥617.8 billion.

By product, capacitors generated ¥282.5 billion, up 30.0% year on year; inductors and EMI filters generated ¥64.3 billion, up 22.4%; and functional devices generated ¥30.5 billion, up 23.0%. By contrast, high-frequency and communications products generated ¥86.1 billion, up only 5.0%, while Energy and Power generated ¥34.8 billion, down 2.7%. While capacitors and peripheral components for AI servers are growing, weakness remains in smartphones and power-related applications. In H1, the issue will be whether data-center-related revenue not only exceeds the ¥69.7 billion recorded in Q1 alone but also absorbs declines in other applications.

The full-year forecast has been raised to revenue of ¥2.110 trillion and operating profit of ¥430.0 billion. Assumptions include an exchange rate of ¥155/USD from Q2 onward and ¥156.12/USD for the full year, as well as smartphone component shipments of 1.19 billion units, automobile production of 90.36 million units and PC component shipments of 360 million units. At the next results announcement, we will check Q2 data-center revenue, the order backlog, the book-to-bill ratio and whether demand assumptions by application have been maintained, to determine whether the full-year forecast is supported not only by AI server demand but also by actual shipments and production.