Supermicro Q1 2027 Earnings: Revenue Nearly Quadruples to $19.9B

Supermicro's revenue nearly quadrupled year-over-year to $19.9 billion in its fiscal first quarter of 2027, the company confirmed on November 3, 2026, as non-stop orders for Nvidia Blackwell-based AI servers pushed results far past the company's own guidance and prompted it to raise its full-year sales target to as much as $82 billion.

What happened

Super Micro Computer (Nasdaq: SMCI) reported unaudited, preliminary results for the first quarter of fiscal year 2027 — the three months ended September 30, 2026 — after market close on November 3, 2026, followed by an earnings call and webcast at 5:00 p.m. ET the same day. Net sales came in at $19.9 billion, nearly quadruple the $5.0 billion booked in the same quarter a year ago and up from $11.1 billion in the prior quarter. GAAP diluted earnings per share rose to $1.75, from $0.26 a year earlier, while non-GAAP diluted EPS reached $1.81, up from $0.35. The company's own numbers, filed with the U.S. Securities and Exchange Commission as an exhibit to a Form 8-K, describe the results as preliminary and subject to change once the quarterly financial close is complete.

Founder, president and CEO Charles Liang framed the quarter, in the official results release, as proof that "Supermicro's transformation into a Total AI and IT Datacenter Solution provider continues to accelerate," crediting the company's Data Center Building Block Solutions (DCBBS) approach — which bundles everything from individual servers up to gigawatt-scale data center designs — for "driving significant value for our customers, from gigawatt-scale datacenters down to single-container PODs."

Why revenue nearly quadrupled

The headline number is not a surprise in direction, only in size. Supermicro had already flagged a sharp upside revision on October 20, 2026, when it issued a preliminary business update lifting its Q1 revenue estimate to a range of $18.5 billion–$19.5 billion from original guidance of $14.5 billion–$15.5 billion, and its gross margin estimate to 12.0%–12.5% from 10.4%–10.8%. The company attributed both revisions to "increased demand for AI platforms" and a shift in "customer and product mix." The final reported figure of $19.9 billion came in even above that already-raised range, meaning actual demand outran even the company's own mid-quarter upward revision.

Behind the number is a company that builds and assembles rack-scale systems around Nvidia's GPUs rather than designing silicon itself. As hyperscalers and AI-cloud operators race to deploy Nvidia's Blackwell and Blackwell Ultra GPU platforms, server builders who can ship complete, liquid-cooled racks quickly have captured an outsized share of that spending. Supermicro's balance sheet shows the strain and the opportunity at once: accounts receivable rose to $9.7 billion from $6.1 billion, inventories climbed to $15.1 billion from $12.9 billion, and current deferred revenue — cash collected for systems not yet recognized as sold — jumped to $4.3 billion from $1.6 billion, all signs of a company scrambling to build and ship faster than it can finish recognizing the revenue.

Supermicro's pitch to AI infrastructure buyers has centered on speed of deployment rather than owning the GPU silicon itself. Its Data Center Building Block Solutions (DCBBS) line, launched as a standalone business in 2025, packages everything a customer needs above the chip: server nodes, networking, power shelves, direct-liquid-cooling (DLC) manifolds, and even prefabricated modular data-center shells, sold as pre-validated, pre-integrated blocks rather than components a customer has to assemble and qualify themselves. The pitch is that a buyer can go from a signed order to a running "AI factory" in months rather than the year-plus timelines that custom-built data centers have historically required. That speed-to-deployment model is exactly what gets stress-tested when demand triples in a single quarter, and this quarter's inventory and receivables growth are the accounting footprint of that stress test playing out in real time.

The numbers, quarter by quarter

Operating cash flow was negative $434 million for the quarter — a reminder that rapid growth in a hardware-assembly business consumes cash even while it is highly profitable on paper, because Supermicro has to pay for components and build inventory well before it collects from customers. Capital expenditures and other investments totaled $47 million. The company ended the quarter with $6.9 billion in cash and equivalents, down from $7.5 billion at the end of June 2026, against $8.6 billion in bank debt and convertible notes.

Guidance: a bigger fiscal 2027

Supermicro's outlook for the current quarter, Q2 FY2027 (ending December 31, 2026), calls for net sales of $19.5 billion to $21.5 billion, GAAP diluted EPS of $1.59 to $1.83, and non-GAAP diluted EPS of $1.66 to $1.90. For the full fiscal year 2027, the company now expects net sales between $72.0 billion and $82.0 billion — a figure that, at the midpoint, would be roughly triple fiscal 2026's revenue. The guidance assumes a GAAP tax rate near 21.2% and a diluted share count around 753 million shares.

That kind of full-year range — a $10 billion spread between the low and high end — also signals how much uncertainty Supermicro itself still has about the pace of AI infrastructure buildouts, GPU allocation from Nvidia, and how quickly it can convert backlog into recognized, shipped revenue.

Margins: better, but not back to historic highs

Gross margin recovered to 12.7% GAAP (12.8% non-GAAP) after bottoming at 9.3% a year ago, when Supermicro was selling AI systems at thin margins to win volume and work through an accounting restatement that had rattled investor confidence in 2024 and 2025. The prior quarter's 17.5% GAAP gross margin was unusually high and not repeated; management's commentary ties the quarter-to-quarter swings to "customer and product mix" — essentially, which GPU platforms and which customers (hyperscalers typically negotiate the lowest per-unit margins) dominate shipments in a given quarter.

Operating expenses grew more slowly than revenue, at $477 million GAAP versus $285 million a year ago, letting operating income swell to $2.06 billion from $182 million and pushing the operating margin above 10% for the first time in several quarters. Interest expense of $83 million was largely offset by $54 million in interest income, and the effective tax rate ticked up to 21.4% from 19.3% a year earlier. Stock-based compensation totaled $100 million for the quarter, including $8 million recognized within cost of sales, a cost base that has grown alongside headcount as Supermicro scales manufacturing capacity in San Jose, Taiwan, Malaysia and the Netherlands to keep pace with orders.

Weighted-average diluted shares outstanding were 749 million on a GAAP basis and 764 million on a non-GAAP basis for the quarter, both up from prior periods as the company has periodically issued equity and convertible securities to fund growth — a dynamic that dilutes per-share metrics even as absolute profit dollars climb sharply.

Balance sheet: fast growth, thinner cushion

The quarter's working-capital swings are the clearest sign of how capital-intensive this growth cycle is. Total assets rose to $35.7 billion against total liabilities of $19.7 billion, leaving stockholders' equity of $16.1 billion. Supermicro also paid $64 million in dividends on its Series A Mandatory Convertible Preferred stock and accrued a further $25 million in preferred dividends not yet paid in cash. None of this is unusual for a contract manufacturer riding a demand wave, but it underscores why the company continues to carry $8.6 billion in bank debt and convertible notes even as net income climbs into the billions.

Why it matters: the AI server supply race

Supermicro's results land in the middle of a broader scramble among server makers — including Dell, HPE and ODMs working directly with hyperscalers — to ship Nvidia's latest GPU platforms into new data centers as fast as they can be built and powered. Supermicro has positioned itself as one of the fastest movers on liquid-cooled, rack-scale designs built around Nvidia's Blackwell Ultra generation, a strategy that let it capture outsized revenue growth this quarter even as margins stayed well below the company's pre-2024 historical range. The same AI buildout has been a tailwind across the chip supply chain this earnings season, and AMD's own data-center revenue surged in its latest quarterly results on similar demand. The competition for GPU allocation and rack capacity is also shaping the next round of accelerator launches, including the rivalry detailed in how AMD's MI450 stacks up against Nvidia's upcoming Vera Rubin platform.

What this means for SMCI stock

Supermicro shares have been among the most volatile in the AI hardware trade over the past two years, swinging on both delivery-execution news and on the accounting concerns that led to a delayed 10-K and a short-seller report in 2024. This quarter's beat — revenue above even the already-raised October guidance, EPS multiples higher than a year ago, and a full-year outlook as high as $82 billion — gives bulls a concrete data point that the AI server backlog is converting into real, recognized revenue rather than just orders on paper. The company's cash burn and reliance on debt and preferred equity to fund working capital remain the main counterargument for skeptics watching the stock, and search interest in "SMCI stock" terms has stayed elevated through results day as retail and institutional investors try to parse whether the margin recovery is durable or a one-quarter mix effect.

Analysts following the stock will be weighing three things against each other in the days after this release: whether the $72–82 billion full-year guidance range is conservative or aggressive given how far Q1 beat its own October update; whether gross margin can climb back toward the mid-teens as higher-margin DCBBS software, networking and service attach rates grow as a share of revenue; and whether operating cash flow turns positive again as deferred revenue from already-collected customer deposits gets recognized in coming quarters. None of those questions can be answered definitively from a single quarter's preliminary numbers, which is also why Supermicro's management continues to caution that the figures in this release are unaudited and subject to change before the formal 10-Q is filed.

Reading the filing yourself

Because these are preliminary, unaudited figures disclosed via an SEC Form 8-K exhibit rather than a completed 10-Q, investors should expect Supermicro to file its actual 10-Q in the coming weeks, which could adjust specific line items even though the headline revenue and EPS figures are unlikely to move materially. The 8-K itself lists customer concentration, declining average selling prices, and tariff exposure among the risk factors management flagged alongside the results. Supermicro has separately disclosed in recent filings that it is building out new manufacturing capacity to handle the order volume, including expanded facilities intended to raise its ability to assemble and liquid-cool complete racks rather than ship servers that customers must then integrate on site.

Where Supermicro sits against Dell and HPE

Supermicro, the San Jose, California-based server maker founded in 1993, is not the only company benefiting from the AI buildout, but it has positioned itself as one of the fastest to bring new Nvidia GPU generations to market in rack-scale, liquid-cooled form. Dell Technologies and HPE both run large AI-server businesses of their own, built on long-standing enterprise relationships and broader product lines spanning storage, networking and services beyond pure AI infrastructure. Supermicro's bet is that a narrower focus on building-block server and rack design, paired with in-house manufacturing it controls directly rather than outsourcing to contract manufacturers, lets it move from a new Nvidia chip announcement to shipping volume faster than larger, more diversified rivals. This quarter's results are the clearest evidence yet that the bet is working at scale, at least for as long as Nvidia's own GPU supply keeps growing fast enough to feed every server builder competing for it.

What happens next

The next milestones to watch are Supermicro's formal 10-Q filing confirming these preliminary numbers, any commentary from Nvidia's own upcoming earnings about Blackwell Ultra supply and allocation across its server partners, and whether Supermicro's Q2 FY2027 print in February 2027 lands inside the $19.5–$21.5 billion range it has now guided to. A full-year outcome anywhere near the top of the $72–$82 billion guidance range would mark one of the fastest revenue run-ups of any publicly traded hardware company in the current AI buildout cycle.

Written byMara Lindqvist

Mara Lindqvist edits the hardware desk. She covers graphics cards, processors, memory and storage, the foundries and chip designers behind them, and what the numbers on a spec sheet mean for people choosing a PC. Specifications in her stories come from manufacturer spec pages and datasheets.

More from Mara LindqvistAll Hardware coverage