美超微 (SMCI) 2027 財年第一季財報電話會議：營收達 199 億美元，上修展望

重點摘要

• 美超微 (SMCI) 公布 2027 財年第一季營收達 199 億美元，年增 298%、季增 79%。其中約有 23 億美元 來自原定於較早時期進行的部署，使得基本營收約為 176 億美元。
• AI 解決方案約占營收的 71%，高於第四季的 59%。管理層預期 AI 相關解決方案將占 2027 財年營收的 70% 至 80%。
• Non-GAAP 毛利率達 12.8%，較去年同期上升 330 個基點，但低於第四季的 17.6%，主因是大型 AI 部署比重較高。
• 該公司將 2027 財年營收展望從 650 億至 720 億美元上調至 720 億至 820 億美元，並預測全年毛利率為 11.8%-12.2%。
• 營運資金效率大幅改善。現金循環週期從 149 天降至 97 天，庫存周轉天數降至 74 天，應收帳款週轉天數降至 36 天。
• 管理層指出，零組件供應狀況、客戶資料中心準備就緒程度、電力、土地和資料中心外殼建築為 2027 財年成長的主要限制因素。

核心財務業績

業務與營運表現

大型 AI 基礎設施部署推動了本季的表現。單一大型資料中心客戶即占 營收的 44%，而 2026 財年全年則為 28%。兩家客戶分別占季末應收帳款的 49% 與 17%。

依地區劃分，美國貢獻了 營收的 80%，其次為歐洲的 11%、亞洲的 6% 以及世界其他地區的 3%。美國營收年增 771%、季增 103%。

管理層將資料中心模組化解決方案（Data Center Building Block Solution，簡稱 DCBBS）定位為成長與毛利率的核心驅動力。該產品整合了機櫃級硬體、液冷技術、管理軟體與部署服務。美超微亦透過與思科 (Cisco)、NetApp 及 Nutanix 的合作，擴大在企業市場的影響力。

公司表示正出貨配備其 DCBBS 與 DRC2 液冷架構的輝達 (NVIDIA) Vera Rubin NVL72 機櫃。公司亦提及 AMD MI350 與 MI355 系統的成長動能，以及支援代理型 AI (agentic AI) 工作負載的 CPU 基礎設施需求走強。

美超微正在擴充矽谷與台灣的產能。管理層表示，公司預計在 12 月前實現每月生產超過 8,000 套機櫃，其中包括超過 4,000 套直接液冷機櫃。

營運資金指標在本季有所改善。庫存周轉天數減少 45 天至 74 天，應收帳款週轉天數改善 23 天至 36 天。遞延營收增加約 31 億美元，主要反映客戶在交付前的付款。應付帳款週轉天數從 29 天降至 13 天，消耗了約 30 億美元的營運資金。

管理層展望

對於 2027 財年第二季，管理層預估：

• 營收為 195 億至 215 億美元
• 毛利率為 12.0%-12.5%
• GAAP 稀釋每股盈餘為 1.59 至 1.83 美元
• Non-GAAP 稀釋每股盈餘為 1.66 至 1.90 美元
• GAAP 營業費用約為 4.55 億美元，其中包括約 1.02 億美元的股權獎勵費用
• 其他淨費用（包含利息費用）約為 5,800 萬美元
• 資本支出為 7,000 萬至 8,000 萬美元

針對 2027 財年，美超微將營收展望從 650 億至 720 億美元上調至 720 億至 820 億美元。公司預計全年毛利率為 11.8%-12.2%。

管理層指出，年度預測假設基本需求將持續呈單季成長，但未假設第一季 23 億美元的追趕性營收會再次出現。毛利率展望隱含下半年的毛利率將低於上半年，因為預計 AI 解決方案在營收組合中的占比將趨近預期 70%-80% 區間的高點。

風險與關注焦點

• 供應限制：管理層表示，需求已超過 GPU 及其他零組件等關鍵組件的供應能力。
• 資料中心準備就緒程度：土地、電力與已完工的資料中心外殼建築仍是限制客戶部署與營收認列時程的因素。
• 客戶集中度：單一客戶即貢獻了季度營收的 44%，而兩家客戶占據了應收帳款的很大一部分。
• 毛利率組合：與更具多元化的企業與通路業務相比，大型 AI 部署的毛利率較低。
• 營運資金需求：庫存增加至 151 億美元以支援預期營收，而供應商付款期限縮短則對現金流構成壓力。
• 總體經濟不確定性：管理層將總體經濟環境列為維持審慎年度營收預測區間的原因之一。

分析師問答亮點

針對客戶預付款項，財務長 Mark Long 表示，信用良好的大型客戶付款迅速，而更廣泛的客戶群則提供更多預付款。管理層預期這種嚴謹做法將繼續支援營運資金效率。

針對毛利率，管理層表示業務已奠定約 12% 毛利率水準的基礎。預估各客戶及產品類別內的毛利率將保持穩定或有所改善，儘管大型 AI 部署占比上升將拉低整體毛利率表現。

針對延遲部署，管理層表示第一季包含較常態更大量的先前延遲營收認列。預期未來不會再出現同等規模的延遲部署認列。

針對資金需求，公司表示目前沒有增資計劃。管理層引用 69 億美元現金、營運資金效率改善以及 18 億美元的應收帳款融資額度，作為財務彈性的來源。

針對更強的營收能見度，管理層指出了創紀錄的積壓訂單、客戶預測改善、地理區域多元化以及企業業務的成長。

法人說明會完整逐字稿

完整財報電話會議逐字稿

管理層陳述

Operator

Thank you for standing by. My name is Jen, and I will be your conference operator today. At this time, I would like to welcome everyone to Super Micro's First Quarter Fiscal Year 2027 Earnings Call. Joining us today are Charles Liang, Founder, President and Chief Executive Officer; Mark Long, Chief Financial Officer; former Chief Financial Officer, David Weigand; and Michael Staiger, Senior Vice President of Corporate Development. Thank you. Michael Staiger, you may begin your conference.

Michael Staiger

Thank you, Jen. Good afternoon, and welcome to Super Micro's conference call for the First Quarter of Fiscal 2027. As you heard, with me today are Charles Liang, Founder, Chairman and Chief Executive Officer; Mark Long, Chief Financial Officer; and Former Chief Financial Officer, David Weigand.

By now, you should have received a copy of the press release from the company that was distributed at the close of regular trading and is available on the company's website. Our call is being webcast live on Super Micro's Investor Relations website and will be available for replay until the conference call to discuss our financial results for the second quarter of fiscal '27.

The content of today's call is Super Micro's property. It can't be reproduced or transcribed without our prior written consent. During this call, we may make forward-looking statements based on current expectations, including with respect to our business plans and performance, ongoing market trends and future financial results. These are subject to a number of significant risks and uncertainties, and our actual results may differ materially.

For a discussion of factors that could affect our future financial results and business, please refer to the disclosure in today's earnings release, our most recent Form 10-K and Q and the reports that we may file on Form 8-K with the Securities and Exchange Commission. All of our statements are made as of today, November 3, 2026, and are based on information currently available to us.

Except as required by law, we assume no obligation to update any such statements. During this call, we will discuss non-GAAP financial measures. You can find a reconciliation of these non-GAAP financial measures to GAAP financial measures in our prepared remarks, which is posted on our website.

Our second quarter fiscal '27 quiet period begins at the close of business Friday, December 11, 2026. With that, let me turn the call over to Charles.

Charles Liang

Thank you, Michael, and thank you all for joining us today. Following fiscal 2026 where we nearly doubled our annual revenue. Fiscal '27 is off to a record start. Q1 revenue reached $19.9 billion between the high end of our preannouncement and well ahead of our original outlook of $15 billion. This is a great achievement, and it shows more and more customers prefer our AI and IT total solutions.

Most importantly, our Data Center Building Block Solution, DCBBS, central to this growth, driven deep customer trust and preference, enhancing our profitability and creating sustainable long-term shareholder value. For over 30 years, Super Micro's focus has always been on delivering optimized total solution value to our customers. We will continue leveraging our engineering technology, global resource and manufacturing capacity with high gear to accelerate the customers' deployment.

Moreover, I'm glad that not just large cloud and neocloud, small cloud company, company and organization across enterprise storage, IoT and both country and corporate serving verticals, relying on and adopting Super Micro's total solution every day. We have executed our commitment to drive higher bottom line profitability in Q1. Non-GAAP gross margin expanded by 330 basis points year-over-year to 12.8%. This expansion was driven by healthier customer mix across enterprise and IT accounts, diverse product line build on our system building blocks and a growing revenue contribution from our higher-margin DCBBS mode.

Including our DCBBS hardware, software and unique service stacks. This directly aligned with our strategic roadmap to build a sustainable double-digit gross margin model by delivering superior technology innovation, faster TTM and TTD, faster time to market and faster time to deployment and total solution value proven over more than 2 years in real world customer deployments, DCBBS delivers 4 key values that nontraditional server OEM and overseas CM can match first optimize TTD time to delivery by delivering factory taste plug and play direct scale systems and DCBBS building blocks complete the solution, including data center management software, directly to customer site.

Second, optimize TTO, time-to-online, for shortening life compute readiness from quarters to months and further down to weeks. And third, DCBBS maximize system uptime through building blocks that eliminating [ thermal ] power, software and service bottleneck and number four, lower TCO by slashing energy consumption and facility downtime to accelerate and improve data center profitability, whether our customers are building gigawatt scale AI factory or deploying rapid single container POD solution. Super Micro's DCBBS help build AI infrastructure faster, better and at a lower cost to our customers because optimized hardware alone cannot run an efficient AI data center, managing thousands of [indiscernible] core GPU rack require intelligent real-time software orchestration.

Super cloud composer, SCM, delivers seamless rack scale management and fuller stack hardware facility integration. Super Micro data center manager, SDM, provides complete power and control from individual now up to full building infrastructure for end-to-end hardware life cycle management. And our Super Micro Operation Manager, SOM suite, featuring director and developer console modules enable bare metal model and tokens service with ease for LLM and agentic AI workloads.

Attaching SCM, SDM and SOM directly to our DCBBS portfolio eliminates software fragmentation for our customers. while driving higher value software categories. Furthermore, SCM enhanced [indiscernible] ServiceNow features, AI driven low cost diagnosis, significantly improving data center uptime and giving our customers complete operational confidence.

Equally critical to our total solution advantage is our new expanded global data center service organization. Here, our on-site deployment teams backed by R&D engineers and well-trained post deployment service staff work directly alongside customers' team through service offering to efficiently maximize cluster uptime.

By using automated [indiscernible], our service team continuously monitoring thermal power delivery and node health in real time to solve potential problems before they cause down time. While our dedicated on-site engineer handle deployment validation and immediately hardware maintenance. This proactive high-touch service model builds deeper customer trust, drive high-value recurring service revenue and our highest commitment, bringing a complete peace of mind to our customers through high-quality infrastructure total solutions.

To realize this unmatched solution efficiency, our global manufacturing footprint and scale advancing rapidly. We are actively expanding our physical footprint with major projects underway in Silicon Valley and Taiwan that will add millions of square feet of fully automated production space and for active service bandwidth. Furthermore, we are accelerating our containerized data center production in Asia and U.S.A. together, our global manufacturing and service network put us on track to produce more than 8,000 orders most powerful racks per month, including over 4,000 DLC racks by this December.

On the technology front, Super Micro continues to lead the market across major silicon platforms. We are among the first to market shipping NVIDIA Vera Rubin NVL72 racks with our DCBBS and DRC2 liquid cooled stack. We recently had also launched the NVIDIA Rubin and Vera Rubin CPU rack, and they are ready to ship now as well. By the way, our optimized [indiscernible] powered solutions also coming soon. For AMD, MI350 and MI355 platform had good momentum now still as we advance to TUS and MI450 total solutions. Additionally, we are progressing Intel Xeon 6+ platform alongside our new ARM HGI CPU-based solutions optimized specifically for agentic AI market.

At the same time, we continue to expand our comprehensive edge in IoT product portfolio, while consistently growing our storage product line, bringing real-time complete package of inferencing capability directly into the intelligent edge. These new products accelerate our strategic focus on our enterprise customers as the company expanded towards scaling agentic model.

Enterprise data governance and workload. We recently announced several major strategic partnerships combining Super Micro's DCBBS, which awarded leading technology company, including Cisco, NetApp, Nutanix and others. With addition any Tier 1 partnership coming soon, we are embedding Super Micro DCBBS footprint deeply into the world's most trusted enterprise. I know some investors and market experts have recently raised concern about AI capital expenditure efficiency and return on investment. Let me be clear. Super Micro is uniquely positioned to thrive in this environment. For over 30 years, we have been focusing on the server and storage industry right here in the United States longer than almost all of our competitors.

Building a mature, well established domestic manufacturing and engineering footprint that our competitors simply hard to match. This U.S. heritage and combined with our time to delivery TTO, time to online, and computing uptime and energy efficiency. Whether our customers building massive Tier 1 AI factory or deploying enterprise agentic AI workload, Super Micro ensures their capital investment convert into revenue faster and more costly efficiently than anyone else, we believe.

We are driving strong internal discipline through continuous operational optimization, factory automation and scale development through technology process refinement and training. We have successfully reduced our inventory and production turn along time from about 100 days to 74 days over the past 3 quarters. Looking ahead, our target is to further streamline our inventory and production cycle down to 50 days over the coming quarters. It's not easy, but we believe we will achieve it.

This ongoing efficiency optimization will continue to strengthen Super Micro's cash flow, working capital cycle and drive higher bottom line profitability. As a matter how competitive the market gets, more customers will adopt Super Micro DCBBS that include data center end-to-end hardware management software and end-to-end data center services.

Now let's turn to our outlook for fiscal year 2027. While our overall demand and our growth momentum remained very strong across every front, we must be recognized that key component shortage continue to impact delivery time line across the IT industry, even GPU. A new customer program and strategic ecosystem partnership are ramping strongly, but we want to remain prudent about some key components, availability and volatile pricing based on what we see today.

We anticipate our fiscal 2027 revenue target will be in the range of $72 billion to $82 billion. That is about 12% higher than the target we set one quarter ago. Before I turn the call over, I want to highlight a few important leadership update. First, we are pleased to welcome a long time industry technology leader, Shesha Krishnapura to our Board of Directors, bringing deep data center operational experience and expertise to support our theory.

Second, my sincere thanks to David Weigand for his 6 years of CFO leadership during our period of massive growth. We look forward to his ongoing support and contribution in optimizing our operation to a much bigger scale. Finally, I'm decided to introduce Mark Long new CFO, whose dedicated, I guess, 30, 40 years, of tech and financial leadership will be a strong path for our next phase of growth. Now I will turn the call to Mark.

Mark Long

Thank you, Charles, and good afternoon, everyone. This is my first call with you, so let me take 2 minutes before we get to the numbers. I've spent much of my career in technology infrastructure finance, including as Chief Financial Officer of a public company in this industry. What brought me to Super Micro is that the hardest part is already built, deep engineering advantage, a real lead in liquid cooling and rack scale integration and a key position in the largest infrastructure build-out of our careers. .

What the company needs now is financial capability that matches its engineering and operating capability. I'm excited to join Super Micro at this pivotal time and look forward to helping Charles and the team continue to deliver an expanding portfolio of innovative solutions to a growing global customer base, including some of the largest, most demanding customers who are shaping the future of the AI era.

To do that will require continued financial discipline, transparency, deeper engagement with stakeholders. -- working capital management and capital structure optimization, which will be my areas of focus. Now to the quarter. Total revenue was $19.9 billion, up 298% year-over-year and 79% sequentially. That is above our guidance range of $14.5 billion to $15.5 billion and above the preliminary range of $18.5 billion to $19.5 billion we provided on October 20.

As Charles noted, approximately $2.3 billion represented deployments originally scheduled for earlier periods due to customer and data center readiness. Excluding that, underlying revenue was approximately $17.6 billion. AI solutions contributed approximately 71% of revenue compared with 59% in Q4. Based on our current backlog, we expect AI-related solutions to represent approximately 70% to 80% of revenue for fiscal 2027. Enterprise and channel revenue was $4.9 billion or 25% of revenue, up 229% year-over-year and down 12% sequentially. OEM appliance and large data center revenue was $14.7 billion or 74%, up 32% year-over-year and 167% sequentially driven by production ramps on several large AI data center design wins.

The 5G telco and edge segment, while still small in relative terms, grew significantly both year-over-year and sequentially. By geography, the United States represented 80% of revenue; Europe, 11%; and Asia, 6%; and the rest of the world, 3%. The United States grew 771% year-over-year and 103% sequential. While the U.S. remains the largest market for large-scale AI data centers, we are seeing growing momentum in other regions as large customers expand globally and governments and enterprises plan and fund significant AI data centers in Europe, Asia and the Middle East.

In the quarter, one large data center customer represented 44% of revenue. compared with 28% for fiscal 2026 as a whole. In our 10-Q, you will also see that 2 customers accounted for 49% and 17% of accounts receivable at quarter end. I want to provide more context for this. Concentration at this level is a consequence of the scale at which AI infrastructure is now being built. A single deployment today can exceed our total quarterly revenue from just 1 year ago. We expect concentration to stay elevated while these build-outs run and to moderate as our enterprise and sovereign businesses grow.

What matters alongside the concentration is the quality of the exposure. So let me do that.

Deferred revenue increased by approximately $3.1 billion during the quarter, an overwhelming majority of which reflects customers paying advance of delivery. Days sales outstanding improved from 59 days to 36 days. We recorded no material credit losses in the quarter. On large deployments, we frequently require deposits and often build against delivering milestones. We evaluate counterparty credit before committing manufacturing capacity and critical components, and we maintained a $1.8 billion receivables purchase facility that gives us flexibility in both cash timing and credit exposure.

We have declining business in this market on terms we were not comfortable with. And in the current environment, I expect we will continue to make that choice. Non-GAAP gross margin was 12.8%, that is 200 to 240 basis points above our guidance range of 10.4% to 10.8% and 330 basis points above the first quarter of last year. On a GAAP basis, gross margin was 12.7%. It is also down 480 basis points from the 17.6% we reported in Q4. And I want to take that apart because a single factor explanation would be misleading.

The key factor impacting gross margin was the Q1 AI revenue mix which increased from 59% in Q4 to 71% in Q1. The customer and product mix within our Q1 business was weighted towards large AI deployments, which carry lower gross margins than the more diversified enterprise and channel business. Benefits from tariff relief and lower inventory reserves during the quarter were largely offset by certain lower-margin deferred revenues we were recognized during the quarter.

I want to be precise about 1 point because I know how the mix figures can be read. The sequential decline is not evidence that AI platform revenue lacks margin. It reflects which customers took delivery in which order. The clearest way to see the underlying trend is to look at the last 2 quarters together, which neutralizes the deferral timing. Across Q4 and Q1 combined, revenue was approximately $31.1 billion and non-GAAP gross margin was approximately 14.5%. For the same 2 quarters a year earlier, revenue was approximately $10.8 billion, and gross margin was approximately 9.5%.

On that normalized basis, revenue nearly tripled and gross margin improved by roughly 500 basis. We continue to demonstrate significant operating leverage. Revenue grew 298% year-over-year against an 89% increase in non-GAAP operating expenses. GAAP operating expenses were $477 million, up 67% year-over-year and 5% sequentially. Non-GAAP operating expenses were $384 million, up 89% year-over-year and 8% sequentially.

The sequential increases reflect higher headcount costs and research and development spending as we invest in multiple new platforms for AI data center and enterprise applications. Non-GAAP operating margin was 10.9% compared with 14.3% in Q4. Here, too, the 2-quarter view is more useful across Q4 and Q1 combined non-GAAP operating expenses were approximately $741 million or about 2.4% of revenue, and non-GAAP operating margin was approximately 12.1% compared with roughly 5.4% for the same 2 quarters a year earlier.

We intend to keep investing in R&D, field service and a global footprint, and we expect operating expenses to keep growing well below the rate of revenue. Other income and expense was a net expense of $34 million, consisting of $49 million of interest expense and other expense offset by $83 million of interest expense on our convertible notes and revolving credit facilities. The tax provision was $435 million on a GAAP basis and $455 million on a non-GAAP basis, with an effective tax rate of 21.4% on both basis.

I would like to expand on some points made in the last earnings call regarding earnings per share because our capital structure now has a moving part that I do not think is well understood. We calculate earnings per share using the 2-class method, which reflects the mandatory convertible preferred stock we issued in Q4. That instrument carries a fixed dividend of approximately $75 million per quarter and converts to common stock in June 2029.

Between now and then, both the dividend and the allocation of earnings to that participating security to produce earnings available to common shareholders this quarter by approximately $312 million in aggregate. GAAP diluted earnings per share were $1.75 against guidance of $0.89 to $0.98. Non-GAAP diluted earnings per share were $1.81 against guidance of $1.01 to $1.10. GAAP basic earnings per share were $1.92.

Now to working capital, which I think will help explain key elements of how we operate our business. Our cash conversion cycle improved from 149 days to 97 days, a 52-day improvement in a single quarter. At our Q1 revenue run rate, that difference represents roughly $10 billion of working capital the business would otherwise have needed to fund. The cash flow statement shows this. We used $434 million of cash in operations on $19.9 billion of revenue. For all of fiscal 2026, we used [ $6.8 ] billion on $39.1 billion of revenue.

Capital expenditure in the quarter was $47 million, so free cash flow was negative $481 million against negative $7 billion for fiscal 2026. Let me break down the 52 days, including both the positives and the areas for potential improvement. Days of inventory improved 45 days to 74 days. Days sales outstanding improved 23 days to 36 days, helped by faster collections and by customer prepayments. Days payable outstanding moved against us from 29 days to 13, days which consumed roughly $3 billion and reflects timing of payments to suppliers in a supply-constrained environment.

The largest single driver was customer prepayment. Deferred revenue increased approximately $3.1 billion in the quarter. Accounts receivable grew $3.6 billion and inventory grew $2.5 billion to support the revenue ahead of us, and prepayments funded most of that build. We expect the cash conversion cycle for the current quarter to be within the band of plus or minus 10% versus Q1, which enables us to appropriately manage our working capital while also executing against attractive market opportunities.

We ended the quarter with $6.9 billion of cash and cash equivalents. Bank borrowing and convertible notes totaled $8.6 billion for net debt of $1.7 billion compared with $1.2 billion at the end of Q4. Inventory was $15.1 billion, up from $12.9 billion. Built to support the revenue we expect in coming quarters. We are compliance with all covenants at quarter end.

On capital structure more broadly, we are starting from a strong foundation. The $5.6 billion we raised in June was raised specifically to fund the working capital behind the orders we are now delivering. As we identify high growth -- high quality growth opportunities, we will optimize our capital structure to support them, and we will do that in ways that fortify the balance sheet without taking on unnecessary costs. That means matching the duration and the form of our funding to the working capital and the assets it supports and being deliberate about when and whether we access markets.

I will be transparent with you every quarter about dilution, maturities and available capacity. We ended the quarter with a record backlog, providing stronger visibility for the current quarter and the remainder of the fiscal year. Turning to guidance. For the second quarter of fiscal 2027, we expect net sales of $19.5 billion to $21.5 billion and gross margin of 12% to 12.5%. We expect GAAP diluted earnings per share of $1.59 to $1.83 and non-GAAP diluted earnings per share of $1.66 to $1.90. That guidance assumes GAAP operating expenses of approximately $455 million, including approximately $102 million of stock-based compensation, which is excluded from non-GAAP.

We expect other income and expense, including interest expense, to result in a net expense of approximately $58 million. GAAP and non-GAAP tax rates of approximately [ 21% ] and GAAP diluted share count of approximately 753 million and non-GAAP diluted share count of approximately 769 million, preferred dividends of approximately $75 million and capital expenditure of $70 million to $80 million. For the full fiscal year, we are raising our revenue outlook to $72 billion to $82 billion from $65 billion to $72 billion.

We expect full year gross margin of 11.8% to 12.2%. Let me connect that to the quarter we just reported. Q1 included approximately $2.3 billion of catch-up from prior period deployments. So underlying revenue was approximately $17.6 billion. The full year range does not assume that catch-up repeats. It does assume continued sequential growth in underlying demand, and it reflects materially better visibility into customer deployment schedules than we had 90 days ago.

Our business continues to strengthen and momentum is building for fiscal 2027. Although we are gated by supply constraints and data center readiness, and we are focused on disciplined execution, creditworthy counterparties and expanding our higher-margin revenue streams, on margin, I want to be straightforward about what the full year range implies. With Q1 at 12.8% and Q2 guided to 12% to 12.5%, the full year range implies gross margin in the second half below the first half. That is a mix effect.

We expect AI solutions to move toward the upper end of the 70% to 80% range as the large deployments in our backlog shipped and those carry lower gross margin than enterprise and channel verticals. Within each category, we expect margins to hold or improve. What governs the range is not demand, and it is not our ability to build. It is the pace at which customer facilities become ready to receive what we've already manufactured and the availability of key composites.

The finding constraints on AI build-outs across this industry right now, are land, power and data center shells. We have better visibility into that than we did, and we will update you as the year progresses. Let me close where I started. This was a quarter of strong growth with high-quality customers. Revenue nearly quadrupled year-over-year. Gross margin came in well above our guidance, and customer prepayments funded most of the working capital needed to deliver in.

Behind those results is a broad and innovative portfolio from DCBBS rack scale and liquid cool systems to enterprise storage, edge and software that positions us for the most demanding workloads and the widest range of use case in the AI building. My focus going forward is the one I described at the outset, financial discipline so that growth is profitable growth, transparency so that you can see how this business works from quarter to quarter, including the timing effects that moves our numbers.

Deeper engagement with all of our stakeholders, working capital management, building on the improvement in our cash conversion cycle this quarter and a capital structure optimized to support high-quality growth from a strong foundation. I'm looking forward to meeting many of you over the coming months. Michael, we are ready for questions.

Michael Staiger

Great. Jen, let's poll for questions. .

Operator

[Operator Instructions] Your first question comes from the line of Katherine Murphy with Goldman Sachs.

分析師問答

Katherine Murphy

First, I want to ask just about the raised revenue outlook. I understand that land, power and data center shells are the binding constraints here. But I wanted to understand in your full year outlook and then in your pipeline, what share of those projects do you have line of sight into the offtakers and the committed volume there? And then I have a follow-up question.

Charles Liang

Yes. I mean, as you know, the power and even GPU some component shortage, right? But because our solution outperformed the industry solution, especially with our DCBBS. DCBBS Not just provide data center hardware building block, software and service, help customer build the data center quicker and also keep their computing power, uptime much better. So indeed, we continue to gain more customers and gain more share. So we feel pretty optimistic for continued market share growth.

Katherine Murphy

And then as a follow-up, could you talk more about the customer prepayments in the quarter? And how we should think about that continuing to support your working capital efficiency going forward? Is that something we can expect in the value that DCBBS and the market share gain strategy that you've talked about?

Charles Liang

Yes. As you know, I mean, now cash flow a little bit tied to total company by the Super Micro because we have a better total solution, again, our DCBBS that attract a lot of customers. So basically, we see a good signS. Most of our customers are able to pay us on time. And we feel very started to continue to grow our business. Mark, do you want to add anything.

Mark Long

Yes, I think the advantage, as Charles pointed out, is really around the quality of the customer base. So very large creditworthy customers who pay very quickly. They provide very strong cash flow for us. But the broad customer base is providing more prepayments. And that discipline enables us to fund a lot more of this working capital. And I think you should expect that to continue.

So we feel like we're in a very good position. The company has done a lot since they raised capital in the summer. And since the cash conversion cycle has improved. So working capital efficiency has gone up substantially.

Charles Liang

Yes. We improve our business in terms of cash flow in 2 areas. One is we improve inventory and production turnaround time from about 100 days to now 70 days, and we believe we will be able to continue to improve that production and inventory turnaround time. Second phase, second point, as Mark just said, we get in a higher-quality customer, and those customers pay on time, or even pay quick turnaround. So we are very solid in this area. .

Operator

Your next question comes from the line of Ananda Baruah with Loop Capital.

Ananda Baruah

Congrats on the progress and Mark, welcome. I have -- yes, 2, if I could. I guess, the first one, Mark, is just on the gross margin and the way to think about the dynamics that are layering in through the balance of the fiscal year. Correct me if I'm wrong here. I actually have to hit the midpoint of your guide while down in the second half of the year, sort of flat around 12% in the next 3 quarters.

So correct me if that's wrong. And then really, the question is, you talked about AI mix sort of being a factor in there. What -- could you share a little bit about inside of the AI customer base, like what are the moving parts on the margins? I think Charles in the past has talked about improvement in AI margins sort of as in the go forward and just interested to know what the dynamics at play are, are there as you go through the rest of the fiscal year inside the AI margins? And then I have a quick follow-up.

Charles Liang

Yes. As I just shared, in last 3 quarters or 4 quarters, our production time and also our production quality, [indiscernible] dramatically improved, and that make our turnaround time quicker and make the customer happier because of beta TPD and better TPO help customer time to online. So customers appreciate that, plus our DCBBS software and service that help customer further. And better quality customer, we continue to gain more customers, especially including the enterprise agentic customers. So all those are positive impacting our business.

Mark Long

Yes. And I think just to kind of build on what Charles said, in terms of the guide for the year, what you're seeing is we have good line of sight. And now the business is stabilizing and kind of has this foundation in this 12% gross margin range with room as we continue to diversify our customer base and our product sets to expand those gross margins further. .

So the precise modeling, you're relatively close on your original question. But for us, the visibility is very strong. And as Charles pointed out, the gross margin across all of our customer categories, hasn't improved. That's really the important thing. And this is a function of what Charles was talking about in terms of our great engineering, our great product set and the execution of the team. So we feel like we're in a very good position at this scale, and we see enormous opportunities going forward.

Charles Liang

yes, like our DCBBS, we provide not just data center building product solution hardware components, but also software also help customer kind of keep up time. So all those almost become one of our business moats. So customers like our product -- more and more customers like our solution.

Ananda Baruah

That's really great context. And the quick follow-up is what's the reasonable expectation for opportunity with the CPU business, just with all the structural tailwinds as you go through the fiscal year here.

Charles Liang

Yes. As you know, agentic AI is getting very popular. So we're getting a lots of customers, including a lots of our enterprise customers, they are buying our CPU-based solution, including our CPU based AMD, including NVIDIA Vera, our NVIDIA Vera solution is ready and including ARM solution. So we see a much stronger CPU demand now and looking forward.

Operator

Your next question comes from the line of Joseph Cardoso with JPMorgan.

Joseph Cardoso

And Mark looking forward to working with you. Maybe just for my first, if I could. It sounds like in your prepared remarks, you're not including any of more catch-up or deferred deliveries in the guidance. But I guess maybe first question here is just, can you just clarify if there are still deferred deliveries that are remaining and just not embedded in the outlook? And then the second portion of that question is, as we kind of strip that out of the numbers, you're still kind of implying a moderation in growth coming into the second half of the year versus the first half. So maybe can you just walk us through the dynamics that are influencing that. And if it has anything to do maybe with the concentration with your largest customer here may be fading going into the back half? And then I do have a follow-up.

Mark Long

Yes. I think the key point about deferred revenue historically kind of looking back, was that there was this set of deliveries and deployments that have been delayed, and they have been carried over from prior quarters. In a typical quarter, you will have some revenue recognition based on deliveries and you may have 1 or 2 deployments that are delayed. But those were -- that was a larger than typical amount of revenue that had been deferred. And when that was released, that caused a number of things.

Revenue went up. And as we mentioned, that was margin dilutive on an overall basis. But going forward, we don't expect the same types and the same magnitude of delayed deployments through the deferred revenue. And most of the deferred revenue is either customer prepayments or we have our service and software revenue that we typically defer over a number of periods.

Charles Liang

Yes. The good thing is since we introduced DCBBS, some of our customers and some of our even our internal people are little bit confused, So that's why they kind of have a certain argument about when to recognize. But now our DCBBS is a much mature business. So not just our internal people are now familiar with our DCBBS and customers understand which services they buy, which services they did not buy. So we should not offer them. So since it's getting much clearer, much mature. So that will have our deferred revenue program.

Mark Long

Yes. And the second question regarding guidance implying moderation in growth in the second half, what you're really hearing is just our sort of range of expectations based on the supply constraints and trying to gauge where the data center readiness for large deployments will land us. But as we mentioned in the prepared remarks, demand remains extremely strong and especially the interest in the DCBBS product portfolio.

Joseph Cardoso

Got it. Very clear. And then maybe as my second one, just dovetailing off of that a little bit, and maybe this is a 2 parter as well. But you're talking about maybe shortages here going into -- or kind of the supply tightness, I guess, in terms of the total ecosystem. Can you maybe just touch on what's the most binding constraint as you think about the 2Q and second half outlook?

Is it largely associated with the GPUs? Or is it other elements on the component side? And then the second question is just more around, obviously, within this backdrop of shortages, pricing has obviously been a strong element in terms of pass-through. Any way you can kind of help us think through as we think about your '27 outlook, how much of the growth is being driven by units as opposed to cost pass-through?

Charles Liang

Very good question. I mean the reason why we have a conservative number for annual revenue, $72 billion to $82 billion because we are concerning about some key components shortage and also macro economy. So if the component shortage won't be a big issue, and if the macroeconomy is healthy then for sure, our second half will be much bigger than our first half. So we hope so, and that's why we are very optimistic to see our second half have a much better performance than the first half. And if that's the case, $72 billion to $82 billion becomes very, very conservative.

Mark Long

Yes. And I think your point about the growth in ASPs really driven by the cost of key components that has been going up sometimes more than quarterly -- our outlook does not -- for the second half does not encompass significant growth in ASPs. So again, we have a relatively conservative approach there. And we're really looking at the details of customer deployments, what they need, which product sets, what our production capacity is and kind of how to manage into that. And as Charles pointed out, our demand far outstrips the availability of key components.

Operator

Your next question comes from the line of Ruplu Bhattacharya with Bank of America.

Ruplu Bhattacharya

Congrats on the role, and I look forward to working with you. Maybe I'll ask you my first question. As you look out over the next 12 months, do you think Super Micro has enough cash and cash flow to self-fund the investments that are needed to support the strong growth that you're seeing, what would cause you to raise more capital? And do you have a preference for debt versus equity? And I have a follow-up.

Charles Liang

Very good question. As you know, by end of the last quarter, our cash is $6 billion. We never have so much cash in the bank. So indeed, the cash flow at this moment, we don't worry. Even in the second half continue to grow strongly if the macroeconomy is healthy. Our cash flow is in much strong position than before ever. .

Mark Long

Yes, as Charles pointed out, I mean starting with $6.9 billion of cash puts us in a very good position. As we mentioned, our working capital efficiency has gone up, our ability to fund our working capital needs through our focus on high-quality customers and some customer prepayments.

That puts us in a great position so that we can optimize our capital structure. We can think about trade-offs if we have very attractive opportunities that require incremental funding we have our $1.8 billion working capital facility, and we have the opportunity to access other markets. As you point out, both the credit and equity markets, at this point, we don't have any plans to consider equity raises. And as far as the debt markets, I think we would look at kind of what is available and how it would best position the company for continued success and provide us with the right leverage structure.

Operator

Due to time constraints, we'll be taking one final question. This question comes from the line of George Notter with Wolfe Research.

Unknown Analyst

It's Bernan Rogers on for George. I guess just a question on what's changed in your conversations with customers in the last 90 days in order to get this better visibility. Like is this a feature of today's market, like customers are extending the forecast time lines that they're giving you? Or is this sort of more of a shift in how you're managing the business?

Charles Liang

Yes. As we mentioned, we have a strong back order. And so we have a choice to select customers and customer order quality, customer quality is very important to us. So we have a chance to selective from some point of view. .

Mark Long

Yes. And I think as Charles pointed out, we are getting better visibility from a number of our customers, some very large, some moderately large, and we're getting visibility across geographies and...

Charles Liang

Enterprises.

Mark Long

Yes. And the growth of the enterprise business is another factor that will provide enhanced visibility and stability to our overall revenue stream.

Operator

We have reached the end of the Q&A session. This concludes today's call. Thank you for attending. You may now disconnect.