Sanmina Corporation (SANM)

Executive Summary / Key Takeaways

• Transformational Acquisition: The $1.6 billion ZT Systems deal (plus $450 million contingent) immediately positions Sanmina as a top-tier AI data center infrastructure provider, with management targeting a $5.7 billion annual run rate and a credible path to doubling revenue to $16 billion by 2027—two years ahead of the original timeline.

• Margin Expansion Engine: The Components, Products and Services (CPS) segment, at 14.6% gross margins, is growing faster than the lower-margin Integrated Manufacturing Solutions (IMS) business (7.7% margins), creating a powerful mix shift that should drive operating margins toward the 6-7% long-term target as AI-related content increases.

• Strategic Geographic Positioning: The India joint venture with Reliance, combined with ZT's U.S. and European footprint, creates a regional manufacturing triangle that directly addresses geopolitical supply chain concerns and positions Sanmina to capture the $500+ billion global data center investment cycle.

• Execution Risk Concentration: With 52% of revenue from the top 10 customers and one customer representing 10.1% of sales, the ZT integration must be flawless. The $2 billion working capital target and two-month Q1 2026 contribution create near-term operational leverage that could amplify any missteps.

• Attractive Risk/Reward Valuation: At $158.09 per share, Sanmina trades at 18.6x free cash flow and 1.0x sales, a discount to AI-exposed peers like Celestica (CLS) (3.3x sales) while offering superior balance sheet strength (0.14 debt/equity) and a clear earnings acceleration story.

Setting the Scene: From Component Supplier to AI Infrastructure Architect

Sanmina Corporation, co-founded by Jure Sola in 1980 and headquartered in San Jose, California, has spent four decades evolving from a printed circuit board manufacturer into a sophisticated provider of integrated manufacturing solutions. The company operates through two distinct segments that tell the story of its strategic direction: Integrated Manufacturing Solutions (IMS), which generated $6.51 billion in fiscal 2025 revenue at 7.7% gross margins, and Components, Products and Services (CPS), which contributed $1.62 billion at 14.6% gross margins. This 80/20 revenue split masks a critical inflection point—the higher-margin CPS business is growing faster and represents the future of Sanmina's profitability.

For at least a decade, Sanmina has manufactured racks for hyperscalers, giving it deep institutional knowledge of data center infrastructure requirements. This experience created the foundation for its most ambitious move yet: the October 2025 acquisition of ZT Systems' data center infrastructure manufacturing business from AMD (AMD). The $1.6 billion deal (comprising $1.46 billion cash and $150 million stock) plus up to $450 million in contingent consideration transforms Sanmina from a contract manufacturer into an original design manufacturer (ODM) with end-to-end system integration capabilities. Management explicitly stated the goal is to "double Sanmina's net revenue within the next 3 years," but recent commentary suggests this could happen in two years, implying a $16 billion revenue run rate by calendar 2027.

The industry backdrop makes this timing critical. Global data center investments are forecast to exceed $500 billion, potentially reaching $800 billion to $1 trillion by 2028, driven by AI workload demands. Sanmina's existing capabilities in liquid cooling racks, custom optical modules, and high-performance storage platforms position it to capture premium value in this cycle. Unlike traditional electronics manufacturing services (EMS) competitors focused on commoditized assembly, Sanmina is vertically integrating into higher-value components and system-level design, a strategy that directly addresses the AI market's need for specialized, mission-critical infrastructure.

Technology, Products, and Strategic Differentiation: The End-to-End Moat

Sanmina's competitive advantage lies in its ability to provide complete solutions from initial design through full system integration, a capability that becomes exponentially more valuable in AI infrastructure where thermal management, power delivery, and interconnect density create complex engineering challenges. The ZT acquisition adds critical ODM expertise, enabling Sanmina to offer "system architecture, electrical mechanical design" for AI workloads, particularly around liquid cooling solutions that are becoming mandatory for next-generation accelerators.

The CPS segment houses the technological crown jewels. Advanced printed circuit boards for defense applications are "doing well and expanding," with management expecting "at least a few great years" ahead. The mechanical business, especially data center racks with liquid cooling, is performing strongly and receiving incremental investment. Precision machining for military applications, optical pluggable modules, and custom memory solutions from Viking Technology all contribute to a vertically integrated component portfolio that competitors cannot easily replicate. This matters because it allows Sanmina to capture margin at multiple levels of the value chain while ensuring supply security for customers—a critical differentiator when component shortages can derail multi-million dollar data center deployments.

The India joint venture with Reliance Strategic Business Ventures, established in October 2022, complements this technology stack by providing a regional manufacturing hub for AI growth. RSBVL acquired 50.10% of Sanmina SCI India Private Limited for $216 million, yet Sanmina maintains controlling financial interest and consolidation. This structure gives Sanmina local market access while sharing capital costs, positioning it to capture India's emerging data center AI opportunities. A new factory coming online early next year will expand these capabilities, creating a geographic hedge against U.S.-China trade tensions.

What makes this differentiation durable is the cumulative investment in regulatory certifications and customer-specific qualifications. In defense, medical, and aerospace markets—where Sanmina has deep roots—product approval cycles can span years, creating switching costs that protect margins. The ZT acquisition brings AMD as a strategic partner, with $150 million in Sanmina stock aligning interests and providing a direct channel to one of AI's key silicon providers. This ecosystem approach contrasts sharply with competitors like Flex (FLEX) and Jabil (JBL), who remain more generalist in their market approach.

Financial Performance & Segment Dynamics: Evidence of Strategic Execution

Fiscal 2025 results validate Sanmina's strategy, with total revenue of $8.13 billion growing 7.4% year-over-year and non-GAAP operating margin expanding 30 basis points to 5.7%. The IMS segment's 7.96% growth to $6.51 billion was driven by "broad-based demand across the majority of our end markets with particular strength in the communication networks and cloud and AI end markets." This broad-based strength is crucial—it demonstrates that the core business isn't being sacrificed for the AI pivot but rather is being pulled along by AI demand.

The CPS segment's performance tells a more compelling margin story. While revenue grew 5.88% to $1.62 billion, gross margins expanded from 12.8% to 13.9% (or 14.6% based on segment table data), driven by "higher revenue, favorable mix and ongoing operational efficiencies." Management has explicitly targeted "above 15%" margins for this business, and Q4 2025's 14.5% non-GAAP gross margin (up 90 basis points year-over-year) shows clear progress. This 600-700 basis point margin premium over IMS is the engine that will drive consolidated operating margins toward the 6-7% long-term target.

Cash flow generation provides the financial flexibility to execute this transformation. Fiscal 2025 produced $621 million in operating cash flow, enabling $114 million in share repurchases while maintaining a fortress balance sheet with 0.14 debt-to-equity ratio.

The company has $239 million remaining on its authorization, signaling management's view that the stock is undervalued. More importantly, this cash generation funds the $85 million in annual capex (approximately 2% of revenue) needed to expand capacity, with management noting they can "ship an additional 30% based on adding people and equipment, depending on mix and plant."

The segment mix shift has profound implications for future profitability. As AI infrastructure requires more custom components—liquid cooling systems, high-density interconnects, specialized power delivery—the CPS segment should grow faster than IMS, pulling blended margins up. Management's commentary that "providing end-to-end solutions for data center AI and other markets, including critical components, will allow the company to deliver better margins than historical" directly supports this thesis. The ZT acquisition accelerates this trend by adding system-level integration revenue that sits above component sales in the value chain.

Outlook, Management Guidance, and Execution Risk

Management's guidance for Q1 2026 reveals the scale of transformation: total revenue of $2.9-3.2 billion includes $850 million to $1.05 billion from ZT Systems for just two months post-close. Annualizing the midpoint suggests a $5.7 billion run rate, confirming the "check the box" on the $5-6 billion annual target. The legacy Sanmina business is guided to $2.05-2.15 billion, representing 4.7% year-over-year growth at the midpoint—solid for a business of this scale but clearly the slower-growth foundation.

The full-year 2026 outlook for the legacy business calls for "high single-digit" revenue growth, with acceleration expected in the second half as "customer inventories were normalizing across various segments, particularly in communications networks" and "new programs to ramp up." This timing matters because it suggests the ZT integration will coincide with an organic growth uptick, creating potential operating leverage if execution is clean. Management's conservative tone—"I think we're more prudent right now in the guidance what we see with all these dynamics that are happening in today's environment"—indicates they are building in cushion for integration challenges and macro uncertainties.

The long-term target of doubling revenue to $16 billion by 2027 (sooner than the original three-year timeline) implies a compound annual growth rate exceeding 40% when including ZT's contribution. This is extraordinarily ambitious for a company that just grew 7.4% organically. The credibility of this target rests on three pillars: the $500+ billion data center TAM, ZT's existing customer relationships with hyperscalers, and Sanmina's ability to cross-sell its component portfolio into these accounts. Management's statement that "we basically said about 6 months ago or 3 months ago, we think we can double the size of Sanmina within 3 years... we see in front of us with legacy business, plus what we believe that we can grow around ZT plus AI opportunities that are in the pipeline that we can accomplish that hopefully, in the next 2 years" frames this as a pipeline-driven opportunity rather than wishful thinking.

Execution risk centers on the $2 billion working capital target for ZT. Jonathan Faust noted, "we spent a lot of time evaluating that inventory position, a lot of discussions with AMD and ZT... there's always risk. There are some risks there, but our intent is to fully evaluate that and both AMD and ZT are committed to that as well." This is the critical variable—if ZT's inventory turns out to be excess or misaligned with customer demand, it could create a cash drain that temporarily pushes net leverage above the 2.0x target range. The company is targeting 1.0-2.0x net leverage over time with investment-grade aspirations, so any working capital build will be closely watched.

Risks and Asymmetries: What Could Break the Thesis

Customer concentration represents the most material risk to the investment case. With 52% of revenue from the top 10 customers and one customer at 10.1%, Sanmina has significant exposure to individual customer decisions. The ZT acquisition likely increases this concentration initially, as hyperscaler customers represent massive revenue pools but also wield enormous pricing power. If a major customer were to in-source manufacturing or shift to a competitor like Flex or Jabil, the revenue impact would be immediate and severe. Management's diversification strategy—growing the CPS segment and expanding geographically through the India JV—directly addresses this, but the risk remains elevated during the integration phase.

The qui tam lawsuit filed in May 2024 alleging False Claims Act violations related to the SCI subsidiary creates legal overhang. While Sanmina is "vigorously defending" the case, the $12 million settlement with Dialight in March 2025 demonstrates that legal disputes can be costly. The environmental liabilities, including a $3 million settlement in August 2024 and $0.6 million criminal complaint settlement in December 2024, while manageable in size, indicate ongoing compliance challenges that could distract management during the critical ZT integration.

Tariff and trade policy uncertainty adds macro risk. Jure Sola's candid admission that "our customers are confused just as much as I am and I'm sure you are too, what's going on" with tariffs reflects the operational complexity of Sanmina's global footprint. While 80% of manufacturing is in foreign operations, providing natural hedging, the company still faces margin pressure if tariffs cannot be passed through to customers. The regional manufacturing strategy—U.S., Europe, and India—mitigates this, but sudden policy shifts could disrupt customer demand patterns.

Technology transition risk in AI infrastructure is asymmetric. If liquid cooling or custom component requirements shift faster than Sanmina's R&D can adapt, the ZT acquisition's value could erode. Conversely, if AI workloads drive even higher density and thermal requirements, Sanmina's integrated design-to-manufacturing capability could command premium pricing. The July 2024 worldwide manufacturing disruption from a misconfigured system update by a network security vendor serves as a reminder that operational excellence is not guaranteed at this scale.

Valuation Context: Pricing a Transformation

At $158.09 per share, Sanmina trades at a market capitalization of $8.78 billion and an enterprise value of $8.21 billion, representing 1.0x trailing revenue and 18.6x free cash flow. These multiples appear reasonable for a business growing organically at 7-8% with 5.2% operating margins, but they do not fully reflect the ZT Systems transformation. The acquisition, closed in October 2025, will materially change the financial profile starting in Q1 2026.

Peer comparisons provide context for the re-rating potential. Celestica, with heavy AI data center exposure, trades at 3.3x sales and 52.9x earnings, reflecting its 16.6% revenue growth and 10.3% operating margins. Flex and Jabil, more diversified EMS players, trade at 0.9x and 0.8x sales respectively, with operating margins of 5.1% and 5.8%. Sanmina's current 1.0x sales multiple sits at the low end of the AI-exposed peer group but above traditional EMS, suggesting the market is beginning to price in the transformation but has not yet fully awarded the premium.

The balance sheet strength supports the valuation. With 0.14 debt-to-equity, Sanmina has the lowest leverage among major peers (Flex: 0.85, Jabil: 2.46, Celestica: 0.39), providing flexibility to fund working capital build for ZT without diluting shareholders. The $239 million remaining share repurchase authorization, while modest relative to the $8.8 billion market cap, signals management's confidence in value creation.

Cash flow metrics are compelling. Price-to-operating-cash-flow of 14.2x and free cash flow yield of 5.4% (inverse of 18.6x P/FCF) compare favorably to Flex (19.6x P/FCF) and Jabil (20.2x P/FCF), while trailing Celestica's exceptional 113.4x P/FCF due to its hypergrowth. The key question is whether ZT's integration will temporarily depress free cash flow through working capital build, as management suggested: "for a period of time after closing, we expect working capital to build to support investment in the growth of the business, which we anticipate will temporarily push our net leverage ratio above 2x."

Conclusion: Execution at Scale

Sanmina stands at an inflection point where a decades-old manufacturing business is being transformed into an AI infrastructure platform. The ZT Systems acquisition provides the technology, customer relationships, and scale to credibly target $16 billion in revenue by 2027, while the CPS segment's margin expansion and the India joint venture provide the profitability and geographic diversification to support this growth.

The investment thesis hinges on execution. Can Sanmina integrate ZT's $5.7 billion revenue run rate without the operational stumbles that often plague large acquisitions? Will the $2 billion working capital target prove to be supported by real customer demand, or will it become a cash drag? Can management maintain customer relationships while reducing concentration risk? The company's strong balance sheet, proven cash generation, and Jure Sola's four decades of leadership provide reasons for confidence, but the scale of transformation is unprecedented in Sanmina's history.

What makes this story attractive is the combination of a massive market opportunity ($500+ billion in data center investment), a clear strategic path (vertical integration from components to systems), and reasonable valuation (18.6x FCF with a 5.4% yield). What makes it fragile is the concentration risk, integration complexity, and the need to maintain operational excellence while scaling at 40%+ rates. For investors, the critical variables are ZT's working capital management and the pace of CPS margin expansion. If both trend positively, Sanmina's current valuation will prove a bargain relative to its AI infrastructure peers. If either falters, the leverage inherent in this transformation could amplify downside risks.