Penguin Solutions, Inc. (PENG) Earnings Call Transcript & Summary
July 7, 2026
Earnings Call Speaker Segments
Suzanne Schmidt
executiveThanks, operator. Good afternoon, everyone, and I appreciate you being with us today for this earnings conference call and webcast, where we'll go over Penguin Solutions' results for the third quarter of fiscal 2026. Joining me on the call are Kash Shaikh, our Chief Executive Officer, and Nate Olmstead, our Chief Financial Officer. The slide presentation that accompanies this call, along with the press release, can be accessed in the investor relations area of our website; we'd urge you to visit that site during the quarter to get the latest information about the company. I also want to remind all of you to review the note regarding the use of forward-looking statements that appears in the press release and the earnings call presentation. Please be aware that on this conference call, the company will make projections and forward-looking statements, which include, but are not limited to, statements regarding the company's growth trajectory, financial outlook, and preliminary expectations for future fiscal periods, business plans and strategy, product development and innovation, market demand and shifts, supply chain conditions and cost assumptions, leadership transitions, strategic agreements, and existing and potential collaborations. Such forward-looking statements rest on current beliefs and assumptions, do not constitute guarantees of future performance, and are subject to risks and uncertainties, including, without limitation, the risks and uncertainties reflected in the press release and the earnings call presentation filed today, as well as in the company's most recent annual and quarterly reports. These forward-looking statements speak only as of the date they are made, and except as required by applicable law, we take no responsibility for publicly updating or revising any forward-looking statements. We'll also be discussing both GAAP and non-GAAP financial measures. Non-GAAP measures should not be viewed in isolation from, as a substitute for, or as superior to our GAAP results. We encourage you to consider all measures when analyzing our performance. A reconciliation of the GAAP to non-GAAP measures is included in today's press release and the accompanying slide presentation. With that, let me now hand the call over to Kash Shaikh, CEO. Kash?
Kash Shaikh
executiveGood afternoon, and thanks for being here on our call covering the third quarter of fiscal 2026. Penguin Solutions just finished a remarkable quarter, posting record results that show how well we've executed and how strongly customers are responding to our AI factory platform strategy. Looking at the company overall, we achieved record net sales along with EPS that came in well above what we had expected. Following such a strong third quarter, we're once again lifting our outlook for the entire year on both net sales and EPS. Every one of our business units turned in a solid performance. The standout growth came from our AI-driven businesses. During Q3, those AI-driven businesses accounted for 74% of total company net sales and expanded 104% versus the same period last year. Integrated Memory and non-hyperscale AI infrastructure solutions make up these businesses. Within them, demand driven by AI kept surpassing net sales growth, which added to a backlog that keeps getting bigger. That gives us confidence the AI opportunity is widening as enterprises increasingly embrace agentic AI workloads powered by inference at scale. What we achieved this quarter shows Penguin Solutions is becoming a leading AI factory platform company. We think we remain in the early innings of a major long-term opportunity for profitable growth. Before I go deeper into our performance, I want to talk about the transition in our finance leadership. As we announced on June 1st, Nate Olmstead will leave the Chief Financial Officer role on July 8th to take an opportunity in another industry. Throughout a significant period of transformation at Penguin Solutions, Nate served as a key partner. His leadership contributed to making our financial and operational foundation stronger, and I want to express my thanks for what he has done and wish him the best. I'm also happy to share that Aaron Johnson, our Vice President of Finance and Accounting, will step in as interim CFO starting July 9th. Aaron carries over 16 years of public company experience in the technology sector along with a deep grasp of our business. We've begun looking for a permanent CFO as well. None of our operating priorities, financial discipline, or focus on execution change because of this transition. Today, before I walk through our third quarter performance, I'll address four key topics. First, the market environment as AI shifts into production scale inference and agentic AI. Second, how our AI factory platform is positioned to capture this opportunity. Third, the customer and partner momentum that confirms our strategy is working. Fourth, how we're executing across product innovation, operations, and go-to-market. Since our earnings call in April, the demand environment has grown stronger as AI keeps moving from early prompt and response experimentation toward production scale inference and agentic AI. An easy way to grasp this shift is that early AI answered questions, whereas agentic AI performs work. During the early phase of AI adoption, a lot of workloads were transactional. Someone asked a question, AI produced an answer, and the session finished. Agentic AI works differently. Agents are persistent, context-rich, and task-oriented. They're able to operate continuously across workflows, applications, and data sources, which moves AI from an advisor into an operator. Take software engineering as an illustration: nowadays an individual developer may supervise several agents handling coding, testing, code review, documentation, and monitoring duties. Over in business operations, agents are able to track what customers are doing, assess risks across the supply chain, get follow-ups ready, and refresh workflow processes. With inference demand growing and agentic AI rollouts speeding up, the infrastructure needed for production AI setups keeps widening over the entire AI data center technology stack. What's rising isn't just demand for GPUs and accelerator-attached high bandwidth memory, or HBM — it also covers general purpose compute, spanning CPU plus memory, storage, and networking. Around every GPU deployment sits a layer of general purpose compute and memory that feeds data, coordinates workflows, handles context, and links agents into enterprise applications. Our view is that this trend is boosting the need for our memory solutions. Such persistent, context-heavy workloads call for greater memory capacity, quicker access to context, and improved orchestration. With AI shifting toward inference at scale, the industry increasingly acknowledges that memory — rather than compute by itself — is turning into one of the main bottlenecks limiting large context AI inference performance. Our MemoryAI appliances were purpose-built to tackle this bottleneck and enhance inference performance along with token economics; in our view, for enterprises, sovereign AI initiatives, and new cloud customers alike, AI factories are turning into a necessity for optimizing time to first token performance and token economics in inference. These shifts bolster the strategic rationale underpinning Penguin Solutions. AI infrastructure at production scale demands a full stack AI factory platform approach. Our platform ties together AI infrastructure, memory, operation software, and operational execution, helping customers deploy and run things at scale. Our second topic today concerns how Penguin is emerging as a leading AI factory platform company at the intersection of memory and AI infrastructure. Five core elements, together with our partner ecosystem, make up our platform. First, ClusterWareAI, the operating system software for AI factories. Second, MemoryAI and integrated memory solutions. Third, advanced computing systems marketed under our Compute AI brand. Fourth, OriginAI factory architectures. Fifth, end-to-end design, build, deploy, and managed services. In our view, what sets our platform apart is the way these proprietary products and services come together as one integrated offering. What customers require increasingly goes beyond merely procuring AI hardware. They require architecture that performs in production at scale, a faster route to deployment, and an operating model that helps them generate profitable revenue or deliver operational efficiencies. In production AI, how long it takes to reach production deployment is directly linked to how long it takes to reach revenue, and our platform is built to shorten both. Customers need the ROI and superior token economics that our Penguin AI factory platform can provide. Moving to our third topic, customer and partner momentum. Through our land and expand go-to-market strategy, we keep adding new customer logos while deepening engagements with existing customers. In Q3 2026, we added four new AI infrastructure customer logos. Securing new logos matters partly because they frequently turn into repeat business. Take the four quarters running from Q3 2025 through Q2 2026: we brought on 13 new AI infrastructure logos, and 7 of those customers have already grown their business with us. Landing new customers may take longer, with extended sales cycles, yet it also builds deeper ties, generates repeat business, and drives more durable long-term growth. Deepgram stands out as a powerful illustration of what our partner model and technical differentiation can do. Penguin engineered and rolled out an optimized inference environment, constructed on Dell infrastructure and NVIDIA technology, to handle enterprise voice AI workloads. During Q3, we broadened our work with Deepgram to add production capacity as its business scales, and Deepgram likewise purchased our ClusterWareAI product along with additional services. Also in Q3, we grew a relationship we had previously disclosed with a tier 1 financial institution. That expansion brings in our Compute Express Link, or CXL-powered MemoryAI KV cache server, ClusterWareAI software, and services, with Dell supplying AI compute. Rollouts like Haein in South Korea, which delivers local Neocloud CPU-as-a-service, keep proving we can meet AI infrastructure needs across many kinds of customer environments. That brings us to our recent Q3 customer win with a leading quantitative trading firm. Spectra, a new sovereign deployment, is another key proof point. This supercomputer system came together and went live via our work with Sandia National Laboratories and NextSilicon. For Penguin, Spectra shows we can design and deploy a complex system within a demanding national security setting.
Kash Shaikh
executiveLooking at our memory segment over the last four quarters, spanning Q3 2025 through Q2 2026, we brought on board 16 new logos, and afterward five of those accounts grew their business with us. We also experienced ongoing expansion with a generative AI client that buys our CXL memory expansion cards to power inference workload solutions. These ties confirm our land and expand approach. When a customer sees the value, performance, and ROI of our AI factory platform, we gain the chance to make our relationships deeper and broader. Penguin's capabilities kept earning recognition from our partner ecosystem. Not long ago we became an NVIDIA AI Factory Specialized Partner, which acknowledges our know-how in designing, building, deploying and managing full stack NVIDIA-based AI factory infrastructure. This quarter also brought recognition as the 2026 Dell Technologies Global Alliances Americas AI Partner of the Year. Now let me offer some perspective on our product operations and go-to-market execution. Within our product portfolio, we keep putting investment into areas where we foresee potential for durable differentiation, such as our MemoryAI products, our ClusterWareAI operating system software, and our OriginAI reference architectures. MemoryAI targets the memory bottleneck that arises as AI inference workloads grow contextually larger, more concurrent, and more latency sensitive. Our CXL-based MemoryAI KV cache server is built to hold large context inference nearer to the workload, letting large language models respond more quickly without having to reprocess whole datasets for each prompt. MemoryAI KV cache can greatly boost AI factory efficiency by unlocking superior token economics for inference and agentic AI workloads, delivering, first, as much as 2x higher inference performance. Second, as much as 8x lower time to first token latency. Third, expanded memory capacity past GPU HBM by using CXL memory that costs roughly 4x-5x less than GPU HBM. The tier 1 financial services customer we added in Q2 bought additional MemoryAI KV cache servers in Q3 for their on-prem AI factory, which initially concentrates on inference and agentic AI for code generation using open weight LLMs. We view AI-assisted code generation as a widespread use case for inference-focused on-premise AI factories, where the proper architecture can yield superior token economics. Our early investment in Celestial AI, a pioneer in Photonic Fabric technology, mirrors our long-held emphasis on memory architecture innovation; we keep advancing our MemoryAI Photonic Memory Appliance, or PMA, through our ongoing partnership with the Celestial AI team, now part of Marvell. This appliance is engineered to extend memory capacity and bandwidth for large context AI inference environments. Taken together, our CXL memory expansion cards, CXL-based MemoryAI KV cache server appliance, and MemoryAI PMA appliance strengthen our innovation agenda and long-term vision for AI memory infrastructure. ClusterWareAI tackles another critical need: managing and orchestrating complex AI infrastructure across heterogeneous environments. ClusterWareAI is a hardware vendor agnostic operating system for AI infrastructure in the data centers. It functions as one consolidated control layer spanning GPUs, CPUs, memory, and networking, which lets the infrastructure run as one cluster, cuts down on manual setup, and helps get to production sooner. Not long ago we rolled out a new AI factory operations agent, giving data center administrators a conversational natural language interface for operational insights. That's the first of a planned family of agents meant to streamline and automate cluster operations and lift administrator productivity, using a human-in-the-loop approach. OriginAI takes these technologies into validated reference architectures, which can lower deployment complexity and risk. Getting deployed faster can translate into faster revenue for our customers. We recently rolled out OriginAI Inference Solutions. Combined, these solutions are built to tap Penguin Solutions' 4 billion hours of GPU runtime experience; they also pull from over 30 years of know-how in delivering advanced memory and AI infrastructure solutions. As our deployments keep scaling across enterprise, sovereign AI, and Neocloud customers, we're centered on boosting repeatability. That covers architectures, software, services, and operations. Over time, this can bolster customer outcomes as well as platform economics. Inside Penguin, we're deploying agentic AI throughout our own operations, centered on measurable business outcomes that back our profitable growth plans; that covers AI-assisted software development, frequently called vibe coding, plus AI-powered productivity tools built to speed up our product cycles, sharpen time to market, and make our supply chain more scalable and responsive; we're also handling supply chain, working capital, and growth investments with discipline as we scale up to satisfy AI-driven demand. Now to our third quarter performance: we posted record quarterly net sales of $479 million, rising 48% year-over-year and 40% sequentially. Non-GAAP operating income hit a third quarter record of $64 million, climbing 67% from the year ago quarter, which shows the strong operating leverage in our model as we scale. AI-driven demand kept fueling growth in Integrated Memory and in the non-hyperscale AI infrastructure business within Advanced Computing. Combined, these two businesses accounted for 74% of total company net sales and expanded 104% year-over-year. Advanced Computing net sales came to $138 million, or 29% of total company, growing 4% year-over-year. Demand and customer engagement for our AI infrastructure business within Advanced Computing stay very strong across enterprise, sovereign AI, and Neocloud customers. Our non-hyperscale AI infrastructure business is scaling quickly, with third quarter net sales climbing 81% year-over-year. In Q3, this non-hyperscale AI infrastructure business made up 58% of total Advanced Computing net sales, versus 33% in the third quarter of last year. The move to inference at scale with agentic AI lines up directly with Penguin's AI factory platform strategy and the differentiated role we play where memory and compute infrastructure meet. Our MemoryAI appliance keeps picking up traction. It produced both revenue and new bookings this quarter. The pipeline also keeps getting stronger across enterprise, sovereign AI, and new cloud customers. Our memory segment posted exceptional net sales of $275 million, representing growth exceeding 111% compared to the same period last year, driven by increases in both units shipped and prices. These impressive outcomes point to a deeper transformation. Agentic AI is fueling persistent, structural need for memory, which strengthens our conviction that demand powered by AI will prove longer-lasting than a typical cyclical rebound in the memory space. We are building our memory operations around the data center sector, where agentic AI serves as the engine behind demand. We are expanding our collaborations with networking, hyperscale infrastructure, and enterprise computing clients regarding their data center offerings. Entering the fourth quarter, we closed Q3 with an exceptionally robust backlog. That speaks to powerful AI-focused data center demand; this demand keeps growing faster than net sales, which gives us a solid backlog as we head into Q4. Equally significant, we are extending that expansion across a broader range of memory customers. This is building a customer base with greater staying power; such momentum underscores how our data center-oriented memory solutions are reaching further. It also brings into focus how substantial the opportunity ahead truly is. Our CXL memory expansion cards keep building momentum, producing revenue as well as fresh bookings during this quarter, while our customer pipeline only grows stronger. Our LED segment likewise delivered solid results this quarter, with Q3 net sales reaching $66 million, an increase of 7% from the prior year. Our strategic focus remains fixed on memory and AI infrastructure as part of Advanced Computing. Meanwhile, LED stays under disciplined management.
Kash Shaikh
executiveIt's producing positive cash flow and keeps moving forward on its innovation roadmap. We made an announcement on June 1st about an improved financial outlook. Back then, our expectation was that net sales and diluted EPS for the full year of fiscal 2026 would land at the high end of the ranges we had previously issued. As of today, with support from strong AI-driven demand and based on what we saw in the third quarter, we're raising our fiscal 2026 outlook once more for both net sales and diluted EPS. Nate is going to walk through more specifics shortly. Going forward, our intention is to keep balancing growth investments against operational discipline. Our priorities are clear. First, put money into differentiated software, AI memory, and compute infrastructure solutions. Second, carry out our plans with discipline and speed. Third, bring in new logos while deepening relationships with existing customers. Fourth, expand our partner ecosystem and make our customer base more diverse. Taken together, these moves are meant to drive more consistent and predictable profitable growth. To wrap up, AI has entered a production phase. That phase is reshaping the architecture and economics of data centers alike. As agentic AI makes inference more persistent and more demanding, a number of capabilities are growing more essential; among them are memory, orchestration, full stack integration, and operational discipline. We think our strategic position sits at the intersection of memory and AI infrastructure, which lets us lead the next phase of AI, inference at scale powering agentic AI workloads; we remain in the early innings of a very large growth opportunity. Thanks go out to our employees, customers, and partners. Our teams feel energized by what's ahead, and we're grateful for their continued focus on execution and customer delivery. With that, I'll hand the call to Nate for a closer look at our financial results and outlook.
Nate Olmstead
executiveThanks, Kash. I'll keep my comments centered on our non-GAAP results, and you can find the reconciliation to GAAP in the tables in our earnings release as well as in the investor materials posted on our website. So with that, let's move on to how we performed in the third quarter. During the quarter, net sales and profits both came in well above what we had anticipated, mainly because demand tied to AI for our memory products kept accelerating and adoption of AI infrastructure solutions continued, and that combination produced a 40% increase in overall net sales from the prior quarter along with a 42% quarter-over-quarter rise in non-GAAP operating income. For Q3 FY 2026, total Penguin Solutions net sales reached a record $479 million, which is 48% higher year-over-year. Non-GAAP gross margin was 28.1%, which beat our expectations and represents a decline of 3.6 percentage points compared with Q3 a year ago. Non-GAAP operating margin reached 13.4%, an improvement of 1.5 percentage points from last year, while non-GAAP diluted earnings per share came to $0.84, which is 79% above year-over-year and 62% above the previous quarter. In the third quarter of fiscal 2026, net sales for our products overall totaled $414 million, which amounts to 87% of total company net sales and reflects 60% growth relative to the prior year. Services net sales came to $65 million, or 13% of total net sales, and declined 1% from the prior year. Here is how net sales broke down by business segment: in Advanced Computing, Q3 net sales reached $138 million, which is 29% of total company net sales, with growth of 4% year-over-year and 19% sequentially. This sales increase reflects stronger AI infrastructure sales to enterprise customers, which more than offset reduced sales to hyperscale customers. Within Advanced Computing, our non-hyperscale AI infrastructure business continues to scale rapidly, with net sales up 81% year-over-year in the quarter. In addition to accelerating growth in this part of our business, we continue to make good progress on diversifying our net sales to new customer categories. In Q3, the non-hyperscale AI infrastructure business represented 58% of total Advanced Computing net sales versus 33% in the third quarter of last year. We continue to see strong demand from enterprise, NeoCloud, and sovereign AI customers and expect these customer categories to represent an increasing share of our Advanced Computing net sales over time; in Integrated Memory, Q3 net sales were $275 million, representing 57% of total company net sales, up 111% year-over-year and 60% sequentially. In Optimized LED, Q3 net sales were $66 million, representing 14% of total company net sales and were up 7% versus the same quarter last year. Non-GAAP gross margin for Penguin Solutions in the third quarter was 28.1%, down 3.6 percentage points year-over-year and down 3.1 percentage points sequentially, primarily attributable to the ongoing wind down of our Penguin Edge business and a shift in the overall mix of sales across our business units. These factors were partially offset by AI-driven demand, which supported favorable pricing in our Integrated Memory business during the quarter. Non-GAAP operating expenses for the third quarter were $70 million, up 9% year-over-year and up 14% versus last quarter. The sequential rise in operating expenses comes from typical seasonal patterns, greater spending on R&D—including for our ClusterWareAI AI software and MemoryAI solutions—and larger variable compensation tied to our robust net sales and profit results. For Q3, non-GAAP operating income reached $64 million, a record for a third quarter, which is 67% higher than a year earlier and 42% above the previous quarter. Operating margins improved by 1.5 percentage points compared with the prior year and by 0.2 points quarter-over-quarter, thanks to solid operating leverage. Non-GAAP diluted earnings per share in the third quarter came to $0.84, representing a 79% gain over Q3 of last year and a 62% gain over the prior quarter. Adjusted EBITDA for the third quarter stood at $68 million, rising 51% year-over-year and 34% versus the prior quarter. Now moving to the balance sheet. On working capital, net accounts receivable reached $704 million, versus $293 million a year earlier, with the growth mainly reflecting substantially larger memory sales volumes and prices. Days sales outstanding stayed healthy at 53 days, an increase from 47 days a year earlier and from 50 days in the prior quarter. Inventory ended the third quarter at $498 million, up from $184 million a year earlier, which reflects higher memory costs, expansion in our memory and AI infrastructure businesses, and strategic buys to secure supply for expected future memory demand. Days of inventory came to 42 days, up from 36 days a year earlier and down from 51 days last quarter, chiefly because of the timing of receipts and shipments. Accounts payable finished the quarter at $736 million, up from $272 million a year earlier, primarily as a result of higher memory costs, growth in our memory and AI infrastructure businesses, and the timing of purchases and payments. Days payable outstanding were 62 days, compared with 53 days last year and 63 days last quarter. Those year-over-year and quarter-over-quarter changes stemmed from the timing of purchases and payments. Even with substantial growth in our net working capital, our cash conversion cycle stayed inside our usual range at 33 days, which is an improvement of five days from Q2 and three days more than last year, owing to the timing of purchases and payments. In line with past practice, days sales outstanding, days payables outstanding, and inventory days are computed on a gross sales and gross cost of goods sold basis, which were $1.22 billion and $1.09 billion, respectively, in the third quarter. As a reminder, the gap between gross and net sales is mostly tied to our memory business' logistics services, which are accounted for on an agent basis, meaning we recognize only the net profit on logistics services as net sales. Cash, cash equivalents, and short-term investments totaled $440 million at the end of the third quarter, down $295 million from Q3 last year and down $49 million sequentially. That year-over-year decline was primarily attributable to proceeds from the issuance of preferred shares in Q2 of last year, offset by debt repayments for our term loan in Q4 of last year. If we take things in order, the main driver of the cash decline was money put into working capital to support our growth, and that was partly offset by roughly $40 million in proceeds from selling our 19% equity stake in Zilia Technologies. Cash used in operating activities during the third quarter came to $75 million, whereas the same quarter a year earlier had generated $97 million in operating cash. The year-over-year drop in quarterly cash flow stemmed mainly from net working capital investments made to back the substantial expansion of our memory and AI infrastructure businesses. If you're among those who follow capital expenditures and depreciation, capex came to $3 million for the quarter, while depreciation totaled $5 million for the quarter. To finish up on cash flow activities, we used $9 million to buy back about 466,000 shares during the third quarter through our stock repurchase program. As of May 29th, 2026, a total of $56 million was still available for buying back our common stock under our existing authorizations. Let's now shift to our outlook.
Nate Olmstead
executiveBecause we delivered strong results through the first nine months and our Q4 forecast for the memory segment has improved, we're lifting our outlook for the whole company's net sales and non-GAAP diluted EPS for the year. At the midpoint, that now implies 22% growth in net sales and $2.60 in non-GAAP diluted EPS, versus the 12% net sales growth and $2.15 in non-GAAP diluted EPS we projected before. Keep in mind, our annual outlook presumes we keep broadening our customer sales mix, and it excludes any Advanced Computing AI hardware sales to hyperscale customers. Also, matching what we assumed last quarter, our FY 2026 financial outlook bakes in the continuing wind down of the high-margin Penguin Edge business; we anticipate sales there will basically stop by the close of fiscal 2026. Together, those two assumptions pull down our FY 2026 growth outlook by roughly 14 percentage points at the total company level year-over-year, and by about 30 percentage points inside Advanced Computing. That said, our full-year net sales outlook carries these full-year growth ranges by segment; for Advanced Computing, we're raising our full-year outlook versus what we previously guided and now anticipate net sales falling 15%-20% year-over-year. As before, this outlook captures the Penguin Edge and hyperscale hardware sales impacts I mentioned. For Memory, we're lifting our full-year outlook, and we now anticipate net sales rising between 90% and 95% year-over-year, fueled by continued AI-related demand and a favorable memory market environment; for Optimized LED, we're likewise raising our full-year outlook and now anticipate net sales dropping approximately 5% year-over-year. Our non-GAAP gross margin outlook for the full year now stands at 28.5% ±0.5 percentage points. We raised our gross margin outlook by half a percentage point to reflect favorable memory pricing, which benefited our Q3 gross margins. Our Q4 outlook assumes less pricing favorability than what we saw in Q3, so we expect some downward pressure on gross margins as we close out the year. Our full-year expectation for total non-GAAP operating expenses has risen to $260 million ±$5 million. For the full year FY 2026, we now anticipate a non-GAAP diluted share count of approximately 56 million shares, higher than our prior outlook, mainly reflecting the expected dilutive impact from our convertible debt due to higher share prices. Because of that dilutive impact, we expect our non-GAAP diluted share count in Q4 FY 2026 to be approximately 62 million shares. Our non-GAAP full-year diluted earnings per share is now expected to be approximately $2.60 ±$0.05. Our forecasted FY 2026 non-GAAP tax rate is now 20%, lower than 22%, reflecting increased pre-tax income in jurisdictions with lower tax rates. Note, our Q3 non-GAAP tax rate was 17.3%, which reflects the year-to-date true-up of this new lower full year tax rate. Although we anticipate applying this normalized non-GAAP tax rate across all of FY 2026 and afterward, over the longer term that rate could shift for numerous reasons — for example, the fast-moving tax landscape in the U.S. and globally, major shifts in the geographic mix of our earnings, or adjustments to our strategy or how we run the business. Because AI-driven demand trends are currently so strong, we'd like to offer some preliminary insight into what we initially expect for net sales and non-GAAP EPS growth in fiscal 2027. Drawing on the customer signals we're seeing now and what we anticipate regarding sustained AI-driven demand, our early fiscal 2027 outlook assumes roughly 30% growth in both total company net sales and non-GAAP EPS, measured against the midpoint of our full year FY 2026 outlook. That EPS projection takes into account the larger share count starting in Q4 2026, which I referenced earlier. This is a first planning estimate; we plan to share a complete FY 2027 outlook during our next earnings call. Our fiscal year 2026 outlook and our preliminary FY 2027 view rest on the present environment and our present assumptions — among other things, assumptions about customer demand, the worldwide macroeconomic backdrop, continuing supply chain constraints, and supply costs, particularly those tied to our Advanced Computing and Integrated Memory businesses. That covers longer lead times for certain components built into our overall solutions, which affects how fast we can ramp both existing and new customer projects and fill customer orders. Our outlook also takes into account the industry-wide rise in memory costs, which could dampen customer demand for our products and solutions and could reduce gross margins in our Advanced Computing and memory businesses. We think the mix of accelerating AI-related demand, a growing enterprise, NeoCloud, and sovereign AI customer base, and our disciplined operating model leaves Penguin well positioned for continued profitable growth. For more detail, please see the non-GAAP financial information section and the reconciliation of GAAP to non-GAAP measures tables in our earnings release and the investor materials posted on our website; with that, operator, we're ready to take questions.
Katherine Murphy
analystThanks for taking that. It was really encouraging to see the Integrated Memory guidance lifted to 90%-95% for the entire year following such a solid quarter. Could you share how much of that comes from the better pricing backdrop as opposed to demand for the new products you mention, such as the CXL cards? Let me get my follow-up in now as well — you also brought up the expanding set of customers for your memory offerings, moving into hyperscale. Could you give more detail on precisely what you're delivering into that market, and what portion of that feeds into the preliminary outlook of 30% growth in total company sales for fiscal 2027? Thanks so much.
Nate Olmstead
executiveThank you, Katherine. Broadly speaking, for our memory business, the improved outlook factors in both the volumes and the pricing; that applies across the business as a whole. CXL makes up a piece of our business. It's a newer product line that's expanding, yet it represents a part of the business, whereas the bulk of the business consists of data center-focused products, and those are the products for which we build specialized memory modules for data center products for our OEM customers. On your second question, one of the customers is a hyperscaler, and we supply a memory module for their data center product for this hyperscaler. It accounts for a portion of the overall business, in the sense of having multiple customers, having multiple products, and it's one of the customers we serve within our memory business.
Kash Shaikh
executiveRight. Kat, I’ll just build on what you said: when it comes to the outlook we provided for FY 2027, this really isn’t a major component of it. We’re not talking about some hyperscale customer worth hundreds of millions of dollars; it’s simply one part of the portfolio, and it’s more or less in line with the run rate we saw in FY 2026.
Katherine Murphy
analystThank you both very much.
Kash Shaikh
executiveThanks.
Brian Chin
analystHello. Good afternoon. Thank you for taking a few of our questions, and all the best to you, Nate. For the first one, I appreciate that you're giving an early fiscal 2027 outlook for total revenue and EPS. I realize you might not want to get too granular yet, but looking at the implied fiscal 4Q 2026, perhaps memory revenue could get to around, say, $300 million per quarter. If I simply hold that flat and it likely rises on pricing and so on, in the coming year that might comfortably be up mid-30%, possibly a higher percent, versus the prior year. On top of roughly that base. So to speak, LED perhaps isn't very growth-oriented in what you're assuming for the coming year. Could you offer some boundaries on how much Advanced Computing might grow in fiscal 2027?
Nate Olmstead
executiveYes. Hey, Brian; this is preliminary, correct? We're in the midst of our own planning for the coming year. We have fairly solid visibility, I'd say, into the first half right now, given backlog and the customer discussions we're having. For Advanced Computing, I'd say probably mid-teens, roughly that, would be sort of where we'd start. Naturally, we'd still have a portion of the year-over-year effect from the Edge business winding down and sort of the move away from Meta, though much of that would already have been eased. What you're beginning to see in the coming year in Advanced Computing is more growth arriving from the non-hyperscale AI infrastructure business.
Brian Chin
analystGreat. That's very helpful. I think Kash and you both touched a bit on the way orders convert to revenue and how Advanced Computing is stretching out somewhat. That involves new customers. It also involves key components, perhaps. I guess, how big a swing factor do you think that could be in fiscal 2027? Are you being pretty conservative about that in terms of anticipating some continued scarcity of memory or other key components to grow those key segments next year?
Kash Shaikh
executiveI think the perspective from the overall business, if we look at our Advanced Computing, especially non-hyperscale AI infrastructure business, some of the advantage we are going to have going into the next year is the bookings for, let's say, AI infrastructure business will deliver the revenues or the net sales going into the first half. That would be the advantage. As we discussed in the last earnings call, our bookings to revenue is, for that part of the business, AI infrastructure business within Advanced Computing, is between three to six months. That obviously provides us an advantage going into the next year; we see at least for the first half, same timeframe in terms of bookings to revenue lag of about three to six months.
Brian Chin
analystOkay, great. Thank you.
Kash Shaikh
executiveThanks, Brian.
Ananda Baruah
analystYeah, appreciate it, everyone. Nice work on everything, and thanks for fielding the question. I suppose just a couple, if that's alright, and I'll put both out there together. Regarding advanced solutions and the widening set of customers, from a services capability standpoint, what kind of engagement are you observing? What are you increasingly delivering for this new, larger customer group? Are you noticing a shift toward more agentic usage? The second one is, as the memory business expands and you're plugging further into the ecosystem with a growing role, is anything being done differently on the business side when you go to market, given it's becoming more strategic in a broader sense? That's all from me. Nate, it's been great collaborating with you once more. I'm sure our paths will cross again down the road.
Nate Olmstead
executiveThank you, Ananda.
Kash Shaikh
executiveWhen it comes to the AI infrastructure segment of Advanced Computing, services give us a strategic edge. Rather than the alternative path, where a customer buys hardware and must piece it together on their own, the benefit we deliver is complete system integration — covering our own offerings plus third-party ones — that encompasses entire AI factories or AI data centers for those clients, spanning design, build, deploy, and manage. That turns into an advantage on our side. In certain instances, we also operate these factories on the customer's behalf for three to five years, which once more benefits the customer by letting them realize the ROI from the solution we provide. Regarding the Integrated Memory business and our strategic direction, one thing we've been doing over the past, I'd say, three months, and increasingly as we move forward, is concentrating it on data centers, correct? It's in data centers that demand is exploding, particularly both accelerated compute, obviously, and now general purpose compute and memory, since agent APIs are fueling that demand. At a high level, our strategy centers on the data center.
Ananda Baruah
analystGreat. Thanks so much.
Kash Shaikh
executiveThank you.
Rustam Kanga
analystExcellent. Thanks to everyone for fielding my questions. Congratulations on those impressive results and the raised guidance. Nate, it's a shame you're leaving, though it certainly seems like you're departing at a peak moment. I'd like to ask about what's being added to ClusterWareAI with the AI factory operations agent. Since these features are increasingly capable of streamlining cluster management and boosting what an administrator can get done, Kash, perhaps you could discuss how ClusterWareAI adoption connects with services uptake, and as these abilities move from a more conversational LLM to more agentic ones, might there be room for additional monetization within that ClusterWareAI side of the business? Thanks.
Kash Shaikh
executiveYes. The AI agent we rolled out is the first member of the agent family. You should view it as clearly helping our services stand apart by delivering software capable of generating automation for both infrastructure deployment and monitoring; that in turn brings them value via our services, and we talk about services and the advantage for our customers when we deliver both product and services. When it comes to building agentic experiences with ClusterWareAI AI, the upside for us is that, in situations where our solution hasn't been fully deployed, whether in a current data center or an AI factory, we're able to give the customer ClusterWareAI AI so they can begin overseeing their current data center or AI factory, and capture the efficiency they might otherwise miss under a traditional deployment; that opens up more opportunity for us to grow our software revenues down the road.
Matt Calitri
analystHi everyone, and thanks for giving me the chance to ask. Nate, I want to add my good wishes; it has truly been a pleasure collaborating with you, and I wish you all the best. What I'd like to know is whether you noticed any shift in the patterns around memory pricing and the supply situation during the past quarter, and whether you can offer any detail on what portion of the quarter's results came from pricing, plus whether enterprise buyers are beginning to push back on the prices at all.
Kash Shaikh
executiveDemand keeps growing. Even though prices are rising, and at some point they might level off, we're still seeing demand grow. As we talked about before, it's a mix of factors — when you look at overall demand across both the backlog and the revenue we managed to ship, demand is climbing and our backlog is climbing too.
Matt Calitri
analystAll right, that's excellent to hear. Did Nate leaving bring about any shift in how you approach guidance? Do you have anything new on the stage of the CFO hunt, or on the kind of traits you'd want in whoever takes over?
Kash Shaikh
executiveTo start with, the transition hasn't altered our operating model or our strategy at all. Our finance and accounting teams are extremely capable, and we're lucky that Aaron Johnson is serving as interim CFO as we conduct the search for a permanent one — he's been at the company, worked alongside Nate, and collaborated with me. That gives the business continuity, and we've engaged an executive search firm for a formal process, reviewing candidates both inside and outside the company. The approach will mainly center on how we maintain the same focus and disciplined execution seen in Q3 and earlier, and how we view the business with real attentiveness to honoring what we've promised. Regarding Nate, there's been no strategic change, and from where I sit the handover has gone quite smoothly.
Nate Olmstead
executiveOh, I absolutely think the handoff to Aaron is going to go really smoothly. For the past 2 years, he's been the person I rely on most, and he's collaborated tightly with me across every part of finance. So my expectation is that this goes very smoothly.
Kash Shaikh
executiveThanks, operator. The results we posted for Q3 show we executed well in a marketplace where demand powered by AI is holding up; the chance that lies in front of us is one we're eager about. Thanks.
Operator
operatorThis concludes today's call. You may now disconnect.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Penguin Solutions, Inc. transcript — plus 132,000+ transcripts from 7,000+ companies, speaker segments, AI summaries and full-text search — through the EarningsCalls.dev API.
Get the API View API docs →This call discussed
For developers and AI pipelines
Programmatic access to Penguin Solutions, Inc. earnings transcripts and 132,000+ others is available through the
EarningsCalls.dev REST API. Plans from $39.99/month — full transcripts, speaker segments,
full-text search, and the recently-added /api/v1/transcripts/recent polling endpoint for ETL pipelines.
More tools for stock research
OptionsBell: an unusual options activity scanner that tracks options flow across all US stocks and emails you when large trades appear in Penguin Solutions, Inc. or any ticker on your watchlist.
StockMarketScan: a stock screener with alerts: 24 pre-built screens refreshed every 5 minutes, watchlists and email notifications. Both offer API and MCP access and fair monthly pricing.