Loading annual review details
Penguin Solutions, Inc. (PENG)FY2026 investor yearly review
The year began with a shift away from hyperscaler-dependent hardware revenue and a back-half growth plan. Stronger memory demand and expanding AI infrastructure wins then accelerated the story, while longer deployment cycles and working-capital demands made execution and cash conversion more consequential.
Overall investor read
Momentum improved through the year
Management described a stronger pipeline and broader customer traction, with acceleration becoming clearest in the second half.
Primary financial signal
$1.73B
FY2026 revenue grew 26.5% year over year, although annual operating cash flow was negative.
Primary call signal
Memory and AI infrastructure scaled
Management increasingly framed the platform around data-center memory, inference workloads, and end-to-end AI infrastructure deployments.
Selected annual financial measures and cash-flow context.
Revenue
+26.5% YoY; second-half acceleration
Revenue growth
Annual growth followed a flat first half
Gross margin
Mix and hardware exposure pressured margin
Operating margin
Annual measure; transcript cited operating leverage
Diluted EPS
+8.3% YoY in supplied annual data
Operating cash flow
Working-capital investment weighed on conversion
Free cash flow
Growth funding left annual cash flow negative
Net debt
Year-end balance-sheet measure
The calls trace a shift from cautious diversification plans to faster AI-led expansion, with memory, customer deployment timing, and cash conversion shaping the quality of growth.
Customer diversification
Management entered the year expecting no hyperscale hardware sales and acknowledged that new-customer deployments could be uneven. Non-hyperscale AI/HPC wins accumulated through the year; by Q4, management described accelerating neocloud demand and more end-to-end platform engagements, while traditional business runoff remained a reported headwind.
Quarter-by-quarter materiality and tone for annual transcript themes.
Scroll horizontally to see more columns.
| Theme | Q1 | Q2 | Q3 | Q4 | Next monitor |
|---|---|---|---|---|---|
| Non-hyperscale AI infrastructure adoption | Mixed New enterprise wins; timing still uncertain | Mixed Seven first-half logos; shipments lag bookings |
AI infrastructure and memory momentum strengthened, but supply, project timing, margin mix, and cash demands remained material execution considerations.
Management cited new customers across enterprise, sovereign AI, and neocloud markets, with later calls describing expanding deployments and bookings. The shift broadened the opportunity set, although management continued to flag lengthy sales-to-revenue cycles.
Management's early discussion of strong demand and CXL qualification became reported deployments, bookings, and record quarterly memory sales by Q4. It increasingly tied the portfolio to data-center and inference requirements.
Revenue was nearly flat year over year in the first half, then accelerated sharply in Q3 and Q4 as AI-driven businesses scaled.
Leverage and cash demands
The story moved from a back-half recovery thesis to visible AI-led acceleration, with memory demand and customer wins gaining scale as execution and cash needs grew.
Management reported modest year-over-year revenue growth despite no hyperscale hardware contribution and framed the year around customer diversification. The outlook depended on second-half deployments, while memory demand was stronger and LED demand remained soft.
Read earnings transcriptMemory demand and pricing led management to raise the full-year company outlook, but Advanced Computing expectations fell as bookings lagged revenue recognition. The new CEO also defined the strategy more explicitly around an AI factory platform and inference-focused memory products.
Only full-year revenue growth is compared; the supplied annual diluted EPS metric is not directly comparable to management's non-GAAP EPS guidance.
22% net sales growth at the midpoint
Revenue 26.5%
The reported annual revenue growth exceeded the midpoint of management's latest FY2026 growth outlook, following two upward revisions.
The next phase hinges on converting expanding demand and backlog into deployments while preserving margin discipline and improving cash conversion.
Long project cycles can shift revenue between quarters and complicate visibility.
Track bookings conversion, customer schedules, and component lead times.
Supply availability and costs affect shipments, mix, and working capital.
Monitor memory sourcing, inventory, pricing, and backlog fulfillment.
Memory and inference
Q1 commentary emphasized strong demand and early CXL qualification. In Q2, management connected inference-focused MemoryAI products to customer demand; Q3 brought reported revenue and bookings for those offerings. By Q4, Integrated Memory was a much larger business, alongside expanding CXL demand and continuing supply and pricing considerations.
Platform execution
Across the calls, management described combining compute, memory, ClusterWare, reference architectures, and deployment services. Product plans became more concrete with MemoryAI and ClusterWareAI updates, while Q3 and Q4 customer examples showed software, services, and infrastructure being included in expanding deployments.
Margins and cash conversion
Management reported sequential operating progress and record operating income as revenue accelerated, while noting mix-related gross-margin pressure. At the same time, calls cited strategic memory purchases and working-capital needs; Q4 management attributed full-year cash use primarily to funding growth in memory and AI infrastructure.
Non-hyperscale infrastructure sales accelerated
Positive Neocloud projects lifted bookings and backlog |
| Track AI infrastructure bookings, deployment timing, customer additions, and conversion into recognized sales. |
| Data-center memory and CXL demand | Positive Memory demand strong; CXL units in early production | Positive MemoryAI launches and CXL customer orders | Positive Memory sales and AI backlog surged | Positive Record memory quarter; backlog exceeded shipments | Watch memory backlog conversion, CXL bookings, pricing, product mix, and supplier availability. |
| Hyperscaler concentration and legacy runoff | Negative No hyperscale hardware; Penguin Edge winding down | Mixed Runoff weighed on sales as new logos grew | Mixed Non-hyperscale growth offset reduced hyperscale sales | Mixed AI infrastructure growth offset traditional runoff | Monitor non-hyperscale mix, traditional-business runoff, and any changes in hyperscale hardware exposure. |
| AI platform software and managed services | Positive ICE software and services framed as differentiators | Positive Platform elements and MemoryAI introduced | Positive ClusterWareAI and inference architectures expanded | Positive Neocloud wins included operation and software | Track ClusterWare adoption, software attachment, managed-services engagements, and repeat customer deployments. |
| Memory supply, pricing, and working capital | Mixed Supply access and memory pricing shaped outlook | Mixed Strategic purchases supported demand amid tight supply | Mixed Inventory investment funded growth and backlog | Mixed Record backlog required substantial inventory investment | Monitor inventory, cash conversion, supplier availability, memory costs, and shipments against backlog. |
| Gross-margin mix and operating leverage | Mixed Favorable mix offset Edge wind-down pressure | Mixed Strong quarter margin; lower second-half mix expected | Mixed Scale improved operating leverage; gross margin fell | Positive Operating margin expanded on scale and expense discipline | Watch gross-margin mix, operating-expense growth, and whether scale sustains operating leverage. |
| Optimized LED performance | Negative China demand weak; some U.S. OEM softness | Mixed Sales remained down as market conditions stayed mixed | Positive LED sales returned to year-over-year growth | Positive Quarterly growth continued; annual sales declined | Track LED demand, profitability, and management's capital-allocation approach to the business. |
The calls moved from describing software, architectures, and services as components of the offer to customer examples combining infrastructure, ClusterWare, and managed services. Management said these engagements can support expansion beyond initial deployments.
As sales accelerated in the second half, management reported record operating income and described expense growth below sales growth for the year. The calls associated the improvement with scale and disciplined spending, while gross-margin mix remained a counterweight.
Management said new-customer contracting, component availability, and deployment schedules can lengthen the booking-to-revenue interval. This uncertainty led to lower Advanced Computing expectations in Q2 even as bookings and new-logo activity improved.
Management repeatedly described tight supply and strategic purchasing to support expected demand. Q4 commentary still linked inventory positioning and shipment cadence to backlog and supply availability, making execution dependent on sourcing and fulfillment.
The calls described increased working capital to support memory and AI infrastructure growth, including strategic inventory purchases. Q4 management attributed full-year operating cash use primarily to those investments, making conversion a separate concern from reported earnings growth.
Management cited lower-margin AI hardware mix, the Penguin Edge wind-down, memory costs, and changes in business mix as pressures on gross-margin expectations. Favorable memory pricing and stronger scale helped at points but did not remove the mix sensitivity.
Scroll horizontally to see more columns.
| Quarter | Revenue | YoY revenue growth | Diluted EPS | Gross margin | Investor read |
|---|---|---|---|---|---|
| Q1Nov-25 | $343.07M | 1.5% | $0.04 | 27.9% | The quarter established a diversification story, but the expected back-half recovery still depended on customer timing. |
| Q2Feb-26 | $343M | -0.0% | $0.58 | 27.3% | Memory supported the revised outlook, but bookings were not converting quickly enough to support the prior AI infrastructure forecast. |
| Q3May-26 | $478.71M | 39.6% | $0.68 | 27.8% | The growth thesis gained reported support, but the cash cost of financing rapid expansion became more visible. |
| Q4Aug-26 | $566.68M | 18.4% | $1.29 | 27.5% | The year closed with stronger backlog and demand visibility, but management still identified deployment cadence and working capital as execution priorities. |
Record quarterly sales and stronger-than-expected earnings shifted the narrative from expected recovery to reported acceleration. Memory and non-hyperscale infrastructure expanded, management raised its annual outlook again, and working-capital investment became a clearer part of the growth story.
Read earnings transcriptThe year ended with record sales, stronger neocloud activity, and management-reported record backlog. Management raised its next-year expectations above the prior preliminary view, while noting that project schedules, component availability, memory costs, and working-capital efficiency could shape results.
Read earnings transcriptNew neocloud programs are increasingly important to infrastructure demand.
Watch new customer wins, project starts, and managed-services attachment.
Working-capital needs can absorb cash even as sales expand.
Track operating cash flow, inventory positioning, and cash conversion commentary.
Hardware mix and memory costs can offset scale-related operating leverage.
Monitor gross-margin mix, operating expenses, and services and software attachment.
Dafinchi