Key Earnings Metrics

Penguin Solutions closed fiscal 2026 with a sharp acceleration in the second half, driven by integrated memory and non-hyperscale AI infrastructure. The company set records across net sales, operating income, and adjusted EBITDA, with operating leverage translating EPS growth well above revenue growth.

• Q4 Net Sales: $567M — Up 68% year-over-year, a company record. AI-driven businesses accounted for 78% of the total and grew 141%, reflecting the shift toward data center memory and AI factory deployments.
• Q4 Diluted EPS: $1.00 — Up 133% year-over-year. Operating expenses rose only 13% in Q4 to $73 million, while full-year OpEx grew 4% to $267 million against 26% revenue growth.
• FY2026 Net Sales: $1.73B — Up approximately 26% year-over-year. Second-half growth reached 58%, compared with a relatively flat first half, marking a meaningful inflection after the AI factory platform launch.
• FY2027 Revenue Guidance: $2.43B — Approximately 40% growth at midpoint, above the preliminary 30% view shared in Q3. Diluted EPS guidance of $4.45 implies EPS growth outpacing revenue growth.
• Integrated Memory Q4: $341M — Up 158% year-over-year and 24% sequentially, representing 60% of company sales. Full-year memory revenue reached $924M, up 99%, driven by higher volume and pricing for data center AI workloads.
• Non-Hyperscale AI Infrastructure Q4 Growth: 99% — Within advanced computing, this segment offset continued runoff from Penguin Edge and lower hyperscale hardware. Advanced computing Q4 revenue was $154M, up 11%.

Management Commentary

Management framed the quarter as evidence that the AI factory platform strategy is working. The company combines ClusterWare AI software, integrated memory, compute systems, factory architecture, and end-to-end services into a single offering. Leadership believes this full-stack approach differentiates Penguin from hardware-only suppliers, particularly for neo-cloud customers that need a single partner to design, deploy, and operate infrastructure.

Inference at scale is creating increasingly memory-intensive workloads, according to management. GPUs remain critical, but CPUs and conventional memory including DRAM must also scale. This dynamic is driving demand across the entire data center stack and supports Penguin's positioning at the intersection of memory and AI infrastructure.

The neo-cloud demand surge reflects strong product market fit, management said. Most projects discussed on the call have already started, with Penguin procuring equipment, building data centers, or operating deployed infrastructure. Engagements typically span three to five years and include ClusterWare AI and managed services with high attachment rates.

On operating leverage, management expects operating expenses to grow considerably more slowly than sales in fiscal 2027. Engineering operations are concentrated in India and Taiwan, manufacturing in Malaysia, and agentic AI tools are being deployed across engineering coding workflows and supply chain functions to compress product cycles.

Memory backlog is higher than revenue recognized in Q4, and the company characterized bookings across both memory and AI infrastructure as strong. Management did not quantify how much of fiscal 2027 advanced computing revenue is covered by backlog, which leaves some visibility gap for that segment.

Strategic Implications

The revenue mix has shifted decisively toward AI-driven businesses. Integrated memory moved from 34% of total revenue in fiscal 2025 to 53% in fiscal 2026, and reached 60% of Q4 sales. This concentration means Penguin's near-term trajectory is tightly linked to data center memory demand and pricing, which management acknowledges is cyclical but argues has a structural AI demand driver.

The advanced computing segment is undergoing its own transition. Non-hyperscale AI infrastructure grew 99% in Q4 and now represents 66% of segment sales, up from 60% a year ago. The wind-down of Penguin Edge and lower hyperscale hardware sales are offsetting some of this growth, but the directional shift toward AI factory deployments is clear.

Penguin's neo-cloud engagements provide a window into how AI compute capacity is being built. The company is designing and deploying an NVIDIA GB-300 NVL-72-based platform for a South Korean tech-backed neo-cloud, and deploying and operating a 36,000-GPU AI factory in Norway for a neo-cloud with $10 billion in contracted compute. These are large, multiyear commitments that extend beyond hardware sales into managed services.

The CXL memory expansion roadmap is another differentiator. Penguin is developing CXL-based large memory expansion cards, a Memory AI KV cache solution, photonic HBM products, and a CXL AIC product to repurpose installed DDR4 memory. These products target the memory bottleneck in AI inference and could expand the addressable market beyond conventional memory sales.

Financially, the company completed a $750 million 0% convertible note offering due 2031 and ended the quarter with $647 million in cash. Capital expenditures remain low at $4 million for the quarter and $12 million for the year, consistent with the asset-light manufacturing model. The new CFO, Stephen Comin, joins as the company enters a phase of rapid scaling.

Margin dynamics warrant attention. Integrated memory economics remain stronger than previously expected, and advanced computing benefits from software and services attachment. However, a higher hardware mix in advanced computing can pressure margins, and management flagged this as a risk. ClusterWare AI and managed services generally carry high attachment rates, which helps offset hardware mix headwinds.

What to Watch Next

I would watch four things over the next two quarters:

• Advanced computing backlog disclosure — The company did not quantify fiscal 2027 advanced computing backlog coverage. If management provides more granularity next quarter, it would reduce uncertainty around the segment's growth trajectory.
• Memory pricing sustainability — Integrated memory growth reflects both volume and pricing. If pricing stabilizes or declines, volume growth from AI workloads will need to carry the segment. Track whether new customer additions and CXL pipeline conversions accelerate.
• Neo-cloud revenue conversion — Several large engagements are in early deployment phases. The pace at which these projects convert from procurement to recognized revenue will shape the quarterly cadence of fiscal 2027.
• Operating expense discipline — Management expects OpEx to grow considerably more slowly than sales. Watch whether Q1 and Q2 operating expense growth stays below 15% year-over-year, which would support the EPS growth guidance of $4.45.

Related coverage

• Base Co. Ltd. FY2026 H1: First Revenue Decline Since Listing, Full-Year Forecast Held
• Accenture Fiscal Q4 2026: $18.7B Revenue, 7% Local Currency Growth, FY2027 Guidance 3-6%
• Micron Fiscal Q4: Revenue Nearly Quadruples to $39.8B DRAM, Q1 Guidance $61.5B
• Paychex Fiscal Q1 2027: Revenue Up 6% to $1.6B, Operating Margin Expands 280 bps to 38%