The Company: Penguin Solutions (NASDAQ: PENG) Overview
Penguin Solutions designs, builds, and manages computing infrastructure for artificial intelligence, high-performance computing, and mission-critical applications. Its portfolio combines GPU computing systems, infrastructure software, deployment services, specialty memory, and fault-tolerant computing. The company also owns Cree LED, which supplies LED components for lighting and display applications.
Financials: Penguin Solutions FY2026 Revenue and Earnings
Penguin’s latest results cover fiscal 2026, ended August 28, 2026, and were released October 6, 2026.
| Financial metric | Fiscal 2026 | Fourth quarter FY2026 |
|---|---|---|
| Revenue | $1.731 billion | $566.7 million |
| Revenue growth, year over year | 26% | 68% |
| GAAP operating income | $165.6 million | $69.5 million |
| Adjusted EBITDA | $256 million | $93 million |
| GAAP diluted EPS | $2.60 | $1.29 |
| Non-GAAP diluted EPS | $2.87 | $1.00 |
Integrated Memory generated $924.0 million, Advanced Computing $558.8 million, and Optimized LED $248.7 million.
Memory drove annual growth, while Advanced Computing revenue declined. Investors should distinguish companywide momentum from AI infrastructure growth. Non-GAAP earnings exclude expenses included in GAAP results.
Bull Case: Why Penguin Solutions Could Benefit From AI Infrastructure Growth
AI deployment complexity creates demand for specialist expertise. Building an AI cluster requires more than purchasing GPUs. Customers must integrate networking, storage, memory, software, and operational support. Penguin’s ability to handle this work can make it valuable to organizations that lack the staff to deploy and operate large systems themselves.
Software and services can strengthen customer relationships. Cluster management, ongoing maintenance, and operational support create opportunities beyond the initial hardware sale. If customers expand their installations, Penguin can pursue follow-on equipment, software, and service business.
MemoryAI addresses a practical AI inference bottleneck. Large language models require substantial memory to retain information used during inference. Penguin’s CXL-based MemoryAI KV Cache Server expands memory capacity outside GPU memory. If the solution improves performance and infrastructure economics in customer deployments, it could become a meaningful differentiator.
Specialty memory offers customization advantages. Industrial, networking, defense, and enterprise customers often need long product lifecycles, qualification support, ruggedization, and dependable supply. These requirements can support closer relationships than those associated with interchangeable consumer components.
The combined portfolio creates cross-selling potential. Penguin can offer customers computing infrastructure and memory expertise within the same organization. The investment opportunity depends on turning these capabilities into sustained customer expansion and profitable growth.
Bear Case: Penguin Solutions Stock Risks and Competitive Challenges
Memory remains cyclical. Demand, component availability, and pricing can change quickly. Strong memory revenue does not necessarily establish a permanent growth rate, particularly when industry shortages or pricing changes influence sales.
Large infrastructure projects produce uneven revenue. GPU cluster deployments can be substantial, but customers may postpone purchases or take longer to complete installations. A successful quarter may therefore be difficult to repeat.
Hardware revenue can grow faster than profit. Expensive GPUs and other purchased components can increase reported sales without producing proportionate margin expansion. The quality of growth depends on Penguin’s integration value, pricing, and software and service attachment.
Larger competitors have substantial advantages. Dell, HPE, and Supermicro can compete through procurement scale, distribution, customer relationships, and broad product portfolios. Penguin must demonstrate that its customization and operational expertise justify the customer’s choice.
Supplier dependence limits control. Penguin relies on external semiconductor and technology suppliers. Product availability, platform transitions, and supplier pricing can affect delivery schedules and profitability.
AI infrastructure spending faces physical and economic constraints. Power availability, cooling capacity, financing, and uncertain returns on AI investments can delay customer projects. Customers may also choose public cloud services instead of owning infrastructure.
Management Outlook: Latest Penguin Solutions Earnings Guidance
Following its latest earnings, management raised fiscal 2027 expectations:
| Fiscal 2027 outlook | Guidance |
|---|---|
| Revenue growth | 40% ±10 percentage points |
| Implied revenue range | Approximately $2.25–$2.60 billion |
| GAAP gross margin | 27% ±2 percentage points |
| Non-GAAP gross margin | 28% ±2 percentage points |
| GAAP diluted EPS | $3.50 ±$0.70 |
| Non-GAAP diluted EPS | $4.45 ±$0.70 |
CEO Kash Shaikh expects continued memory strength and accelerating AI Infrastructure activity. Penguin added 17 AI Infrastructure customers during fiscal 2026.
The implied revenue range is calculated from reported fiscal 2026 sales. Guidance represents management’s expectations, rather than guaranteed results.
TAM / CAGR: Penguin Solutions’ AI Infrastructure Market Opportunity
The broad global AI infrastructure market is estimated to generate approximately $497 billion in spending in 2026, with spending projected to reach approximately $1.21 trillion by 2030. The forecast implies roughly 30% compound annual growth from 2025 through 2030.
This market provides a useful measure of the opportunity surrounding Penguin’s AI computing, memory, and infrastructure services. Growth drivers include AI inference, enterprise deployments, sovereign AI initiatives, and specialized cloud providers.
However, the entire market is not directly available to Penguin. It includes substantial spending by hyperscalers and purchases that Penguin may never supply or manage. Its realistically addressable opportunity is narrower, centered on infrastructure integration, specialty memory, management software, and services.
Penguin also participates in fault-tolerant computing and LED components. These markets overlap with parts of its portfolio but should not simply be added to the AI infrastructure estimate, because doing so could double-count spending.
Products: Penguin Solutions Portfolio and Revenue Breakdown
Penguin discloses revenue by business segment, rather than by individual product. The percentages below use fiscal 2026 revenue and total 100%. Individual product and service revenue shares are not publicly disclosed.
| Business segment | FY2026 revenue share | Major products and services | Applications and competing offerings |
|---|---|---|---|
| Integrated Memory | 53.4% | SMART Modular DRAM modules, including DDR5 products; enterprise, industrial, and ruggedized SSDs; embedded flash storage; CXL memory expansion cards and modules; SMARTsemi discrete memory components | Supplies memory and storage for servers, networking equipment, industrial systems, and embedded applications. Competes with specialty memory suppliers such as Kingston, ATP Electronics, and Apacer, alongside overlapping offerings from semiconductor manufacturers. |
| Advanced Computing | 32.3% | OriginAI infrastructure solutions; GPU and HPC systems; ICE ClusterWare management software; MemoryAI KV Cache Server; Stratus ztC Endurance, ztC Edge, ftServer, and everRun; infrastructure design, deployment, support, and managed services | Supports AI training and inference, technical computing, and applications requiring continuous availability. AI infrastructure competes with Dell AI Factory, HPE Private Cloud AI, and Supermicro AI SuperCluster. Stratus offerings face specialized fault-tolerant systems and high-availability alternatives. |
| Optimized LED | 14.4% | Cree LED XLamp, J Series, and high-brightness LED components | Used in indoor and outdoor lighting, architectural lighting, horticulture, signage, video displays, and specialty illumination. Competing suppliers include Nichia, ams OSRAM, and Lumileds. |
Percentages may sum to 100.1% because of rounding.
Key products within Advanced Computing include:
- OriginAI: Validated AI infrastructure architectures supported by integration and deployment expertise. It helps customers assemble the computing, networking, storage, and software needed for GPU clusters.
- ICE ClusterWare: Software for deploying and managing computing clusters, helping administrators operate infrastructure at scale.
- MemoryAI KV Cache Server: A memory appliance offering up to 11 TB of CXL-based memory for AI inference. It is designed to reduce memory constraints, improve response latency, and limit unnecessary recomputation.
- Stratus ztC Endurance: Fault-tolerant computing for critical applications, using redundancy, proactive monitoring, and predictive failure handling.
- Stratus ztC Edge: Computing designed for industrial and remote environments, with built-in virtualization and simplified management.
- Stratus everRun: Availability software that uses paired servers to protect applications against interruption.
Business Model: How Penguin Solutions Makes Money
Penguin combines product sales with engineering, software, and services.
In Advanced Computing, customers purchase infrastructure and may also pay for design, integration, deployment, management software, maintenance, and ongoing operational support. Initial hardware purchases can generate substantial revenue, while continuing services create opportunities for repeat business.
Integrated Memory sells qualified memory and storage products to equipment manufacturers and other customers with specific performance, reliability, or lifecycle requirements. Customization, testing, and supply support help distinguish these offerings from standard components.
Optimized LED sells components to lighting and display manufacturers through direct relationships and distribution channels. Its economics depend on product performance, manufacturing efficiency, demand, and pricing.
Penguin’s central business proposition is to reduce the difficulty of deploying and operating complex systems. Sustained profitability depends on how much value it captures from engineering, software, and customer support alongside purchased hardware.
Customers: Who Uses Penguin Solutions’ Products?
Penguin serves enterprises, AI companies, specialized cloud providers, government and sovereign AI projects, research organizations, industrial businesses, and equipment manufacturers.
Publicly identified customer examples include:
- Deepgram: Uses infrastructure delivered by Penguin to support speech-to-text, text-to-speech, and voice AI capabilities.
- Lektra: Selected Penguin’s Full-Stack AI Factory Platform for distributed AI data centers serving customer requirements for location, energy availability, and sovereignty.
The broader customer base varies by business. AI infrastructure customers need scalable computing and operational expertise. Memory customers need qualified components and dependable product availability. Stratus customers prioritize application uptime, while Cree LED customers integrate components into finished lighting and display products.
Technology suppliers and channel partners should be distinguished from end customers: a partnership alone does not establish a customer relationship or its revenue contribution.
Competitors: Top 3 Direct Penguin Solutions AI Infrastructure Rivals
The following three companies are major direct competitors to Penguin’s AI infrastructure offerings. They are not a disclosed ranking by competitive market share.
| Competitor | Directly competing products | Competitive strengths and overlap |
|---|---|---|
| Dell Technologies | Dell AI Factory with NVIDIA, PowerEdge GPU servers including XE9680, AI storage, deployment, and support services | Competes in enterprise AI infrastructure and integrated deployments. Dell’s distribution, installed customer base, and breadth are significant advantages. Dell also supplies technology used in some Penguin deployments, making the relationship both competitive and collaborative. |
| Hewlett Packard Enterprise | HPE Private Cloud AI, HPE ProLiant Compute XD systems, AI storage, networking, and professional services | Competes through integrated private AI environments and large computing systems. HPE combines infrastructure with enterprise software and services, addressing customers seeking a complete deployment platform. |
| Super Micro Computer | AI SuperCluster, NVIDIA HGX GPU servers, GB200/GB300 rack-scale systems, and liquid-cooled infrastructure | Competes directly in GPU systems and rack-scale AI deployments. Its broad hardware selection and integrated cooling offerings can appeal to customers prioritizing infrastructure configuration and deployment speed. |
Penguin’s differentiation rests on customization, infrastructure management, memory expertise, and ongoing services. Its memory and LED businesses face separate competitors, as outlined in the product table.
Founding History: From SMART Modular to Penguin Solutions
Penguin Solutions traces its origins to SMART Modular Technologies, founded in 1988 as a specialty memory business. SMART Modular became publicly traded in 1995, followed by a series of ownership changes and reorganizations.
The business completed an initial public offering as SMART Global Holdings in May 2017.
In 2018, SMART Global acquired Penguin Computing, expanding into high-performance computing, AI, and integrated compute and storage solutions. This acquisition established an important foundation for today’s AI infrastructure business.
The company added Stratus Technologies in 2022, broadening its capabilities in fault-tolerant computing and critical application availability.
In October 2024, SMART Global Holdings changed its corporate name to Penguin Solutions. The rebranding reflected its shift from a specialty memory heritage toward a broader portfolio spanning AI infrastructure, advanced memory, software, and services.
