Super Micro Computer, Inc. (SMCI) Earnings

Management highlights

### Strategic Transformation - Transformed from a U.S.-based server manufacturer to a leading AI IT data center total solution company (DCPBS), with manufacturing facilities in the U.S., Taiwan, Malaysia, and the Netherlands. - Positioned as a one-stop shop for customers to build AI data centers/factories faster, with turnkey DCPBS integrating GPU/CPU servers, storage, direct liquid cooling, high-speed networking, data center management software, and full lifecycle services. This cuts customer deployment time from years to quarters, reduces TCO, and accelerates time to revenue. - Expanded DCPBS with a new proactive service model that automatically alerts customers and deploys teams to fix failed units, preventing compute capacity reduction. New software and attached services deepen customer loyalty and long-term value, with more features launching early next quarter. Product and Partnership Highlights - Maintains strong partnerships with top silicon providers: shipping volume NVIDIA GB300, HGX B300, B200, RTX 6000 Pro, while preparing first-to-market new NVIDIA systems; launched a full new AMD Helios product line alongside strong momentum for EPYC CPUs and MI accelerators; launched Intel Parecel Deck Edge AI systems and ships Xeon 6 Plus in volume; developing products for the ARM AGI Conan Phoenix architecture optimized for high-performance inference workloads. Operational and Capacity Expansion - Driving higher manufacturing yields via factory automation, design optimization, and versatile building block architecture, while focusing on logistics and inventory management to reduce inventory reserves and expediting charges. These disciplines moderate quarter-to-quarter margin fluctuation and support consistent gross margin growth. - Expanding global manufacturing footprint: new 32-acre DCPBS campus in Silicon Valley brings total U.S. footprint to nearly 4 million square feet; all global facilities are ramping, with total manufacturing capacity on track to exceed 6,000 racks per month, including over 3,000 direct liquid cooling racks per month. Financial and Customer Highlights - Record full-year fiscal 2026 non-GAAP diluted EPS of $3.63, up 76% year-over-year from $2.06. Non-GAAP operating margin expanded to 8.1% from 7.1% in fiscal 2025. - Customer base is diversifying: 9 customers generated over $1 billion in annual revenue in fiscal 2026, up from 4 in fiscal 2025. One large CSP customer represented 28% of full-year 2026 revenue. - Q4 non-GAAP gross margin hit 17.6%, up 750 basis points sequentially, with 75% of the gain from favorable customer/product mix (deferral of low-margin contracts to future quarters) and 25% from lower tariffs and reduced inventory reserves. Non-GAAP operating margin was 14.3% in Q4, up from 7.2% in Q3. - Completed $5.6 billion in public financing in Q4, leaving a strong balance sheet with $7.5 billion in cash and cash equivalents at quarter end, and no current plans to utilize the previously announced APM program. Net debt dropped to $1.2 billion from $7.5 billion at the end of Q3.

Guidance

- First quarter of fiscal 2027: Revenue is expected to be $15.0 billion to $15.5 billion. Non-GAAP gross margin is projected to be between 10.4% and 10.8%. GAAP diluted EPS is expected to be $0.89 to $0.98, while non-GAAP diluted EPS is projected to be $1.01 to $1.10. Capital expenditures are expected to be $50 million to $60 million. - Full fiscal year 2027: Management reaffirms revenue guidance of $65 billion to $72 billion, driven by record ending backlog of over $60 billion in new orders received in Q4. - Management expects consistent long-term gross margin improvement, driven by balanced customer/product mix between higher-margin enterprise/CPU offerings and lower-margin high-volume GPU solutions, plus growing contributions from higher-margin DCPBS solutions.

Segment performance

For the fourth quarter of fiscal 2026 (ending June 30, 2026): - Total revenue: $11.1 billion, up 93% year-over-year and 9% quarter-over-quarter, at the low end of prior guidance range of $11 billion to $12.5 billion - Enterprise and channel: Revenue of $5.6 billion, representing 50% of total Q4 revenue. This is up 172% year-over-year and 98% quarter-over-quarter. - OEM appliance and large data center: Revenue of $5.5 billion, representing 50% of total Q4 revenue. This is up 50% year-over-year and down 26% quarter-over-quarter. - AI solutions: Contributed approximately 60% of total Q4 revenue (down from over 80% in Q3, due to timing of large AI project ramps; management expects >80% of revenue will be AI-related going forward). For full fiscal year 2026: - Total revenue: $39.1 billion, up 78% year-over-year from $22 billion in fiscal 2025. - Enterprise and channel: Revenue grew 39% year-over-year, representing 31% of total full-year revenue. - OEM appliance and large data center: Revenue grew 104% year-over-year, representing 69% of total full-year revenue. By geography for Q4 fiscal 2026: - U.S.: 71% of revenue, up 259% year-over-year and 12% quarter-over-quarter - Asia: 11% of revenue, down 50% year-over-year and 13% quarter-over-quarter - Europe: 8% of revenue, up 4% year-over-year and 25% quarter-over-quarter - Rest of world: 10% of revenue, up 296% year-over-year and 1% quarter-over-quarter

Risks & headwinds

- Forward-looking statements are subject to material risks and uncertainties that could cause actual results to differ materially from expectations, detailed in the company's SEC filings. - There is inherent inventory risk associated with frequent GPU platform technology transitions; if customer deployment schedules or platform configurations change, the company could be left holding obsolete or excess inventory. - Large AI project deployments can face short-term delays related to customer power, cooling, and data center readiness, which can create quarter-to-quarter revenue and margin volatility. - The company faces competitive pressure from ODMs for large data center customer contracts, and ongoing industry shifts in buying patterns could impact market share.

Analyst Q&A

• Q: An analyst asks what puts and takes drive fiscal 2027 gross margins, how to estimate full-year margins, and what the 20% non-AI revenue segment consists of. He also asks for an update on the ongoing board inquiry. /

  A: Management explains they will balance revenue growth with profitability: high-volume GPU has lower margins, while enterprise CPU, storage, and IoT have higher margins. They will grow enterprise lines and mature DCPBS solutions (which carry better margins) to drive consistent overall margin improvement. Of the ~80% AI-related revenue going forward, 60-70% is pure GPU AI, 10-20% is CPU-based/agentic/edge AI, leaving 20% for traditional non-AI server, storage, and IoT. Management stated an update on the board inquiry will be provided shortly, with no additional comment on the call.

• Q: An analyst asks how the company manages inventory risk amid frequent GPU platform transitions, given past inventory issues, and asks how to think about working capital intensity and the need for future external financing at the projected $65-72 billion revenue scale. /

  A: Management mitigates inventory risk by securing non-cancelable purchase orders from customers, matching procurement to shipment schedules, and leveraging the company's building block architecture that makes most subsystems compatible across multiple product generations, reducing obsolescence. For working capital, management expects the cash conversion cycle to improve in fiscal 2027 due to improved contract terms with customers, and the current strong balance sheet is sufficient to fund projected $65-72 billion growth; incremental external financing would only be considered if revenue grows beyond $80 billion.

• Q: An analyst asks if the company is seeing a shift in AI demand from training to inference workloads, and how this shift impacts CPU/GPU shipment mix. /

  A: Management confirms the overall mix remains ~70%+ GPU and ~high 20% CPU. Inference GPUs carry margins between traditional training GPUs and CPUs. Long-term, the overall GPU percentage will continue to grow as AI inference penetrates all vertical industries, and the company's DCPBS strategy is positioned to maintain planned profit margins through this shift.

• Q: An analyst asks for updates on the DCPBS business's 2026 gross profit contribution and 2027 outlook, and asks about tariff rebates in the strong Q4 gross margin results. /

  A: Management confirms DCPBS (which includes hardware, software, deployment services, and proactive maintenance) has strong long-term growth potential, and a 20% gross profit contribution target is not far out of reach. Management also confirmed no tariff rebate was booked in Q4 results, the company is actively pursuing refunds but will not recognize benefits until they are received.