2026-03-21XQuote

Source item

Smuggling AI servers into China didn’t change the downward trend in Super Micro’s gross margin Excerpt: "...servers sold for $510 million between late April 2025 and mid-May 2025..." Full article: https://t.co/jFjc4ZzU7z At least ~10% of Super Micro’s 2Q CY2025 revenue was tied to servers reportedly smuggled into China. In theory, these should carry much higher margins, but gross margin still declined sharply to 9.6% (vs. 11.3% in 2Q CY2024). Two possibilities: 1. Margins in the legitimate business are simply too weak. This is consistent with my earlier view that AI server assembly margins are under pressure (https://t.co/jofSOlgaEx). Super Micro’s structural disadvantages, including smaller scale and weaker execution, further amplify margin pressure. 2. Super Micro likely wasn’t the only one involved in smuggling. With alternatives available, buyer leverage increased, so margins on those sales were probably not as high as expected.

Quoted context

These are the quoted author’s words.

AI server assembler Wiwynn’s recent 4Q25 gross margin miss (7.2% vs. consensus estimates of 8–8.3%) has triggered a share price correction and reignited investor concerns. While rising component costs can mechanically dilute reported gross margins, it is more critical to examine the underlying profitability trends in server assembly through the lens of structural changes in AI server design. Nvidia continues to increase the level of design integration in AI servers to boost token output per unit of space and power, addressing the constraints of limited data-center space and scarce electricity supply. Highly integrated designs also help improve manufacturing efficiency, which in turn benefits supply-chain management and reduces maintenance costs. VR200 NVL72 is a representative example of this high-integration approach. Based on my industry surveys, its compute tray adopts a cable-less design, resulting in an ~40% reduction in the number of component items vs. GB300 NVL72. However, higher integration also comes with a greater share of Nvidia-specified components and a shrinking scope for customization, both of which are unfavorable to assembler profitability. My industry survey indicates that components classified as “customization not allowed” in the VR200 NVL72 compute tray have risen to 20–22%, well above the 5–7% level seen in GB300 NVL72. Moreover, VR200 NVL72 has directly eliminated "ODM-designed" component items entirely. In fact, this trend toward higher integration began with GB300 NVL72. My industry checks suggest that white-box GB200 NVL72 assembly gross margins were still around 18–20% in 1Q25, but by 1Q26, white-box GB300 NVL72 margins are expected to decline sharply to 6–8%. Importantly, this trend is likely to persist. Looking ahead, Kyber rack designs are expected to feature an even higher degree of integration and face greater margin pressure than the current Oberon generation. Assemblers will not remain passive in the face of these challenges. Beyond straightforward improvements in manufacturing efficiency, there are currently three practical approaches to addressing margin pressure in the server business, although their effectiveness remains to be seen: 1. optimizing product mix by increasing exposure to higher-margin ASIC AI servers; 2. securing NPI opportunities to enhance bargaining power with customers during the early stages of mass production; and 3. pursuing vertical integration strategies—ranging from qualifying in-house or group-affiliated server components through Nvidia validation, to expanding the business scope toward providing full data-center build-out solutions.
mingchikuo ↗