Supermicro earns $19.9B, raises annual forecast to $82B

Super Micro Computer reported revenues of $19.9 billion for the first quarter of fiscal year 2027, ending September 30, nearly quadruple the amount from a year earlier, according to the earnings report released on Tuesday (3) after U.S. market close. The company raised its full-year revenue projection to a range of $72 billion to $82 billion, up from $65 billion to $72 billion announced alongside the fiscal fourth quarter 2026 report. Shares rose about 7% in after-hours trading.

The revenue figure exceeded the preliminary guidance the company had provided on October 20. For the second fiscal quarter, Supermicro is projecting sales between $19.5 billion and $21.5 billion, indicating stability around the $20 billion per quarter mark, rather than a new acceleration.

The margin tells the other half of the story

The adjusted gross margin stood at 12.8% (12.7% on a GAAP basis), down from 17.5% in the prior quarter, according to the earnings release and conference call. The full-year guidance is 11.8% to 12.2%, and for the second quarter, it's 12% to 12.5%. In other words, the company is selling significantly more but retains a smaller slice of every dollar.

Management's explanation is the mix: AI systems, built around GPUs purchased from Nvidia and AMD, carry lower margins than conventional servers, and the company expects AI solutions to represent between 70% and 80% of fiscal year revenue. The integrator passes a costly component to the end customer, whose price it cannot control, and charges for rack engineering, liquid cooling, and delivery.

On the same night, AMD reported data center revenue of $8.5 billion and healthy chipmaker margins. The comparison between the two earnings reports shows where value in the AI supply chain is being captured: in silicon, not assembly.

The skeptical argument, and what weakens it

Debate surrounding Supermicro features three recurring points among analysts: margins, customer concentration, and cash consumption. Operating cash flow was negative for the quarter, and a significant portion of revenue relies on a few large customers, whom the company does not identify. For skeptics, revenue that quadruples while margins decline and cash is consumed appears more a matter of purchased volume than market power.

However, there are data points countering the bubble reading. The company exceeded its own guidance from two weeks prior, raised the annual projection floor by $7 billion, and maintained second-quarter guidance at the same level as the first quarter, rather than promising a jump. A company inflating expectations tends to do the opposite.

The distinction the market tends to overlook is who is picking up the tab. If orders are coming from hyperscalers financing capex with operating profits, demand is more stable than if they stem from AI labs and "neoclouds" that rely on debt and funding rounds. Supermicro's balance sheet does not break down these figures by name, and this absence, more than the margin, should concern investors.

Where the impact appears

In the United States, where Supermicro is based in San Jose, the $20 billion quarterly scale intensifies competition with Dell and HPE for the same AI rack contracts. For American CIOs, this means more competition among integrators, but not necessarily lower prices, as the GPU remains the component defining the cost.

In Taiwan, where the supply chain of boards, chassis, and cooling components that supplies integrators like Supermicro is concentrated, the elevated annual guidance translates into firm orders for local suppliers throughout 2027. In Europe, where banks and governments are discussing AI capacity under data sovereignty rules, the elevated forecast indicates that delivery timelines for complete racks remain contested alongside orders from American clouds.

For corporate buyers, the useful data lies in negotiation. With gross margins around 12%, the integrator has little room to discount hardware. Conversations about the total cost of an AI project increasingly revolve around energy, cooling, and delivery timing, not just server price.