Dell(DELL.US)FY2026 Q4 Earnings Call: AI Revenue Expected to Double to $50 Billion in FY2027

Dell (DELL.US) held a conference call to report its fiscal year 2026 fourth-quarter earnings. The company’s executives stated that the total AI order backlog for FY26 reached $64.1 billion. This quarter, new orders totaled $34.1 billion, with ending backlog at $43 billion, demonstrating strong ongoing momentum. AI revenue for FY27 is expected to double to $50 billion. The customer base has surpassed 4,000 clients (including sovereign clouds, enterprises, etc.), with barriers built through TCO optimization, rapid deployment, and DFS financing.

Executives also noted that high AI demand has caused component supply shortages, rising input costs, and longer delivery cycles. The company has shortened quotation validity periods and strengthened close coordination between supply chain, sales, and pricing to cope with fluctuating costs. Despite significant revenue growth, the company is automating processes with AI, and expects FY27 operating expenses to grow only in the low single digits, reflecting strong economies of scale.

For guidance, Q1 revenue is projected between $34.7 billion and $35.7 billion (a median year-over-year increase of 51%), with AI server revenue expected to contribute $13 billion. Infrastructure Solutions Group (ISG) revenue is forecasted to grow over 100% YoY, while Client Solutions Group (CSG) revenue is expected to grow about 2%. Full-year FY27 revenue is estimated between $138 billion and $142 billion (a median YoY increase of 23%). EPS is expected to be $12.90 ± $0.25, roughly 25% higher YoY.

Q&A

Q: Regarding AI business, orders performed strongly this quarter. In the context of memory price fluctuations, has the profit margin been affected? As the business scales and diversifies its customer base (from CSP to enterprise), what is the potential for profit margin improvement or growth in other areas?

A: This quarter, our AI business performed exceptionally well, with orders reaching $34 billion. More importantly, our pipeline for the next five quarters has not only remained intact but has grown further, covering all customer types including CSPs, sovereign clouds, and enterprise clients. Especially in Q4, enterprise demand was very strong; we now have over 4,000 enterprise AI customers, with significant quarter-over-quarter growth, deploying AI across various enterprise applications.

Regarding profit margins, we have maintained mid-single-digit operating profit margins over the past year and this quarter. Based on current market demand and operational conditions, we see no reason to change this guidance. We will continue to expand our business at this profit margin level.

We are excited about future growth prospects, expecting AI business scale to double next year compared to this year. Last year, orders increased sixfold, driven by explosive inference demand, which boosts token generation. The growth in tokens directly increases demand for compute capacity and intensity, ultimately translating into long-term revenue streams.

Q: Given the explosive growth in AI server orders, has profitability (margin) changed compared to previous quarters? What is the outlook?

A: Regarding AI server margins, we emphasize again that even amid rapid growth, we can maintain operating profit margins in the mid-single digits. Despite undergoing significant technological transformation, we are capable of sustaining this profitability while transitioning. Currently, we have $43 billion in backlog orders, all to be delivered at mid-single-digit margins. With backlog, future demand, and transformation certainty, we are confident in our guidance to maintain profitability.

Q: How has rising memory prices specifically impacted the profitability of traditional server (ISG) and client solutions (CSG) businesses?

A: Our response to rising memory costs has been highly effective:

1. Traditional servers: We responded swiftly. Starting December 10 last year, we adjusted prices for server products. Since then, despite higher costs, margins have stabilized.

2. CSG (PC-related): We adopted a more cautious and targeted pricing strategy. To strengthen market share, we delayed price increases from October to December. This resulted in 18% growth in that quarter, outpacing the 10% market growth, and increased market share by 1%.

Price adjustments: In December, due to increased bidding activity and new customer acquisition, our business composition shifted. We implemented a price increase on January 6, which normalized performance and stabilized margins as expected.

In summary, we are fully capable of operating within our profit framework.

Q: Why has the growth rate of traditional server business not reached a higher (e.g., robust double-digit) level? Also, are inference workloads mainly deployed on AI servers, and how is traditional server inference usage?

A: The demand for traditional servers has exceeded supply, achieving double-digit growth across all customer types and regions. The core driver is the trend of data center “consolidation” and “modernization”: customers upgrading old tech to higher efficiency, e.g., moving from 14G to 17G servers at a 6:1 or 7:1 ratio, bringing significant power, space, and thermal advantages. We expect this modernization trend to continue through FY27.

Additionally, AI inference workloads on x86 traditional servers are increasing significantly. Leading companies, dubbed “AI Forward,” are deploying AI applications on traditional servers for software development, scientific computing, and financial algorithms. This means inference workloads are not confined to dedicated AI servers but are widely distributed across various architectures.

Guidance reflects that Q4 demand far exceeds supply, and Q1 is expected to maintain strong double-digit growth. We are cautiously optimistic about H2, considering supply-demand fluctuations, but if current demand persists, further upside is possible.

Q: After the Vera Rubin cycle, compared to the Blackwell cycle, how do management see the transition’s smoothness, profit volatility, and revenue fluctuations?

A: We are excited about the Vera Rubin transition. Learning from Grace Blackwell’s implementation, we expect Rubin’s transition to be smoother. We’ve applied manufacturing and testing experience to next-gen architectures, enabling faster ramp-up. Early engineering samples are complete, and we’re collaborating with customers on advanced design, with promising progress. Revenue volatility should be manageable due to diversified customer base, backlog, and Rubin’s expected H2 shipments.

On margins, we believe AI can sustain mid-single-digit operating profit margins, even with Rubin’s introduction, based on current backlog and new order outlook. This aligns with our financial guidance. As engineering capabilities improve and engineers deepen customer relationships, transition smoothness will be better than before.

Q: What are the expectations for free cash flow in FY2027?

A: FY26 was very strong, with operating cash flow of $11.2 billion (including $4.7 billion in Q4). This supported share repurchases of over 54 million shares and returning over 80% of adjusted free cash flow. While we do not provide specific cash flow guidance, FY27 is expected to be another strong year. Net profit to adjusted free cash flow conversion should meet or slightly exceed our long-term value creation targets. Today, the company announced a 20% dividend increase, marking the fourth consecutive year of double-digit dividend growth. Coupled with share repurchases, we are confident in future operational stability.

Q: What is the underlying logic for storage business rebound? Under current component (memory) cost increases, is the higher proportion of high-margin storage the key to maintaining overall gross margin?

A: We are very optimistic about storage recovery. This quarter, our proprietary product portfolio grew in double digits, including PowerMax, PowerStore, PowerScale, and data protection platforms. Demand for all-flash storage increased double digits, with new customers across regions. PowerStore has grown for eight consecutive quarters, with seven quarters of double-digit growth; half of PowerStore customers are new, and nearly 30% are new to Dell storage.

Unstructured data storage demand is driven by expanding AI inference and AI applications. Our IP products’ share of total business has increased significantly YoY, and we expect FY27 to outperform FY26, with storage’s share in the portfolio growing further—key to profit contribution.

I believe “architecture” is critical now. For example, PowerStore’s 5:1 data reduction improves capacity; data protection products achieve up to 75:1 compression and deduplication. Amid rising memory costs and supply shortages, these architectures help customers back up data with fewer servers and drives, creating a competitive advantage that will sustain storage growth into FY27.

Q: What assumptions does the company make regarding memory price inflation in FY2027?

A: While specific procurement ratios are confidential, market trends show DRAM spot prices have risen nearly 5.5 times in six months (~$2.39/GB), NAND prices nearly 4 times (~$0.20/GB). Analysts forecast Q2 prices to rise 20–50% QoQ, with continued single- to double-digit increases in Q3 and Q4. Our cost planning incorporates these expectations, with long-term agreements (LTA) and capacity contracts in place. To mitigate cost volatility, we are reducing product complexity, optimizing materials, and designing flexibly to adapt to available components.

Q: What different strategies has the company adopted to protect margins during this cycle compared to previous ones?

A: We have reused best practices learned during the pandemic, executing faster than ever. In traditional servers, we adjusted prices within days of December 10 last year; in PC, we changed quotes for thousands of unshipped orders in a single day on January 6. Previously, we aimed to recover two-thirds of added costs within 90 days; now, responses are immediate.

We leverage multiple measures: raising list prices across categories, introducing intelligent pricing and profit floors, shortening quote validity (now at record lows), and significantly reducing promotions and special pricing.

These strategies have proven effective. Server margins stabilized quickly despite cost increases, showing strong performance. For PC, we delayed price hikes to gain market share (achieving 18% growth vs. 10% market), but since January 6, margins have normalized immediately.

Q: Can you break down the share of enterprise AI server orders among enterprise, emerging cloud providers, and sovereign clouds?

A: We closely monitor the health and growth of emerging cloud, sovereign cloud, and enterprise segments. While exact proportions are confidential, key indicators show strong enterprise momentum: over 4,000 AI customers, record high enterprise revenue in Q4, and expanding AI use models.

Q: When do you expect enterprise AI demand to spill over into storage and trigger a new storage cycle?

A: For example, two years ago, we deployed coding assistants consuming GPU capacity; since mid-last year, we’ve deployed “Agents” that write software based on architect specifications. This has caused a surge in compute demand, especially token consumption. This evolution from simple aids to complex agents is not unique; many leading firms see huge potential and benefits from deploying AI, which will drive enterprise AI volume.

A key signal is that in our pipeline for the next five quarters, enterprise AI growth is the fastest, surpassing other segments. This explosive demand will naturally increase the need for underlying data support, leading to the next storage cycle.

Q: Why has inventory on the balance sheet increased significantly?

A: The inventory increase directly reflects business scale and growth rate. Despite higher inventory levels, our cash conversion cycle remains healthy at -32 days, even shorter by 1 day YoY. This demonstrates disciplined management of working capital amid AI expansion and shipments. We forecast $13 billion in AI shipments for Q1, meaning we are actively managing billions of dollars in equipment shipments in February and March. The current inventory position supports this volume.

Q: How have customer procurement behaviors changed following recent price increases? Has shortening quotation validity caused concerns or preemptive buying?

A: Customer reactions differ between ISG and CSG. For ISG, after initial “price shock,” urgency to secure supply grew quickly. Top-tier, mature clients recognized the seriousness and took aggressive actions to protect infrastructure projects, affecting AI, traditional servers, and storage.

For CSG, the situation differs due to existing low-cost inventory in channels. Cost pressures weren’t immediate, which is why we delayed price increases to gain share. However, large bidding projects in H1 have led customers to accept higher costs and tighter supply. They realize “today’s price is likely better than tomorrow’s,” leading to some preemptive purchasing. While difficult to quantify, this may temporarily deplete budgets and extend replacement cycles. Our guidance accounts for these dynamics with cautious assumptions.

Q: Regarding the $43 billion AI order backlog, what is the split between Grace Blackwell and Vera Rubin? How does this affect future shipment pace?

A: The majority of the $43 billion backlog is composed of Grace Blackwell orders; Vera Rubin is mainly in our pipeline for the next five quarters. The largest portion within this pipeline is Blackwell, especially x86 Blackwell, driven by enterprise deployments focused on air-cooled solutions, scientific computing, and high-frequency trading.

Q: The company set an AI revenue target of $50 billion annually. How is capacity being scaled to meet potentially higher demand?

A: The $50 billion target is based on a comprehensive assessment of deployment progress, infrastructure availability, key component supply (DRAM, E1, E3 drives), and our delivery capacity, made after four weeks into the new fiscal year. We are actively sourcing more components, converting pipeline potential into actual orders, and ensuring timely fulfillment. We will continue to enhance our supply chain agility to meet evolving demand.

Q: Why is there a roughly $5 million difference between the $9.5 billion shipment and $9 billion revenue for AI?

A: The difference is due to normal accounting for “in-transit goods.” About $9.5 billion worth of goods shipped at the end of January, with some not yet delivered to customers, causes revenue recognition to be delayed by a few days, which is typical.

Q: How do you explain the logic behind the single-digit growth guidance for traditional servers, in terms of shipment units and ASPs?

A: Demand for traditional servers is real and expanding. Customers are migrating to 16G and 17G servers with more DRAM and storage. The “consolidation” logic is clear: a 16G server replaces five older units; 17G replaces seven. Customers prefer higher-performance, higher-capacity servers.

For FY27 (calendar 2026), we expect unit shipments to decline, but value (TRUs) to rise due to higher configurations and prices. Our guidance remains cautious for H2, considering supply-demand uncertainties. Q1 demand exceeded supply, showing strong double-digit growth, but we base guidance on current supply conditions. If demand remains strong, we will seek more components globally.

Q: What is the attach rate for AI server shipments? Besides servers, what other products or services are offered to AI customers?

A: Our cross-sell opportunities in AI servers focus on three areas:

• Storage: Enterprise AI growth leads to more storage configurations.

• Networking: Our networking solutions are integral to AI infrastructure.

• Services: Our installation, deployment, and break-fix services are key differentiators. Dell’s ability to deliver complex AI architectures with high uptime and on-site support is unmatched.

These cross-sell rates are rising and well received; we remain optimistic about this growth into FY27.

Q: Regarding the structural share growth in CSG, can it be expected that Dell will surpass the market in FY27 or FY28, leveraging stronger pricing flexibility, demand guidance, and supply chain advantages (compared to smaller competitors)? What are the underlying drivers for share gains amid unit shipment declines?

A: Historically, Dell has excelled during industry-wide shortages, gaining share across all segments, especially PCs. Our deep partnerships and supply agreements give us an advantage. This explains why we didn’t rush price hikes in Q4 but adopted a more aggressive market stance, as we are at a pivotal point to change our trajectory.

After three years of share loss, we reversed the trend. Q4 results show 14% revenue growth and 18% unit growth YoY (per IDC). This momentum is crucial, expanding our customer base and re-establishing market leadership.

We believe that in the coming years, Dell can achieve structural share growth in PCs, driven by supply chain strategies. This advantage extends to servers and storage, and we are confident that with better supply chain execution, we can outperform even in downturns.