San Jose, California, March 16, 2026, 07:30 PDT
Shares of Super Micro Computer climbed roughly 2% to $31.37 Monday morning as Nvidia’s GTC conference got underway in San Jose, putting the AI-server specialist back in focus following a stretch of scrutiny over its ability to maintain breakneck growth. CEO Charles Liang says Supermicro is pushing ahead with rapid scaling for major AI and enterprise rollouts, with its Data Center Building Block Solutions already moving out in significant volume.
The timing is notable: Super Micro has turned into a barometer for AI infrastructure investment. Nvidia’s yearly event is set to highlight inference—the process of deploying trained AI models—plus networking and “AI factory” infrastructure. That’s exactly where Super Micro is pushing past its server roots, aiming instead for full rack and data center projects. Reuters
Super Micro’s DCBBS bundle—servers, power, cooling, networking, software—aims for 6,000 racks a month by the end of fiscal 2026. That target includes 3,000 liquid-cooled racks. The company is already shipping 150-kilowatt racks at scale.
Investors aren’t left guessing. Super Micro posted December-quarter revenue of $12.68 billion, a jump of 123% year over year, and bumped its fiscal 2026 sales outlook to a minimum of $40 billion, up from the previous $36 billion mark. The order backlog? Back in November, the company said it had more than $13 billion worth of Blackwell Ultra systems on the books.
Liang claims DCBBS lets customers boost scale rapidly, cut costs, and go greener. Tech analyst Gadjo Sevilla at eMarketer notes Super Micro’s expansion is closely linked to its status as the main “integrator to large cloud and AI customers”—a spot that often brings in orders well before manufacturing even begins. Super Micro Computer
Rivals aren’t standing still. Dell last month projected AI-server revenue will soar 103% to roughly $50 billion by fiscal 2027. HPE, for its part, said this month that suppliers everywhere are contending with pricier memory, supply constraints, and demands to focus on more profitable orders.
The pressure is clear in Super Micro’s numbers. Gross margin slid to 6.3% for the December quarter, down from 11.8% the previous year. The company blamed a mix of tariffs, pricier production and expedite fees, plus inventory write-downs and a changing mix of products and customers—each eroding profits as new AI GPU systems scaled up.
The balance sheet tells its own tale. As of Dec. 31, receivables hit $11.0 billion and inventories reached $10.6 billion, up sharply from $2.2 billion and $4.7 billion at June’s close. Through the first half of fiscal 2026, operating cash flow — the company’s core cash generation — landed at negative $941 million, mostly due to bigger outlays for inventory and a spike in receivables.
The risks are straightforward. Super Micro might keep landing AI contracts, but high component prices—or sharper price cuts from competitors—could still trip it up. HPE warned that hardware costs might not ease until at least 2027, underscoring that robust demand doesn’t guarantee fatter margins for AI infrastructure players.
Investors are left debating whether Super Micro’s pitch—DCBBS growth and better operations leading to fatter margins—will actually play out, with the market still holding out for hard evidence. The company gets a shot at GTC this week, but product demos alone probably won’t be enough to close the case.
