The company said that “design win upgrades” shifted approximately $1.5 billion in expected first-quarter revenue into the current quarter.
The delay stemmed from a high-volume customer’s request to modify the configuration of graphics processing unit (GPU) racks, Super Micro added.
CEO Charles Liang said the delays were "largely caused by the complexity of these new graphics processing unit racks, which require intricate integration, testing, and validation."
"The profit opportunities have been dramatically different than the revenue opportunities in AI compute, with AI server leaders continuously sacrificing margins to participate in large deals, leaving limited profit upsides for investors to cheer about," J.P.Morgan analysts said.
Super Micro has become a key player in the AI server boom, supplying high-performance systems to data centers racing to scale. Its rapid growth and close ties with Nvidia have attracted investors, though analysts have flagged that the sector's pace has exposed execution risks and margin volatility.
The company's close partnership with Nvidia allows it to be the first to market with fully integrated systems built around new GPU architectures, including Nvidia's Blackwell Ultra series - a key driver of Super Micro's $13 billion GB300 order book, and position itself as a critical supplier in the AI infrastructure race.
"Super Micro's pursuit of lower-margin business and deep ASP discounts to secure GB300 orders were not adequately reflected in the current valuation," analysts at Susquehanna said.
The company raised its full-year revenue forecast to at least $36 billion, up from $33 billion, and projected second-quarter revenue of $10 billion-$11 billion, well above Wall Street's $7.83 billion estimate.
Super Micro has gained nearly 56% so far this year and is trading at a price-to-earnings ratio of 16.94, compared with 9.75 for Hewlett Packard Enterprise and 14.11 for Dell Technologies.
