Penguin Solutions, Inc. (NASDAQ: PENG) – Q2 2026 Earnings
Penguin Solutions, Inc. (NASDAQ: PENG) – Q2 2026 Earnings
Press release and earnings call link
Section 1: Short Tear Sheeet
Penguin Solutions is an AI infrastructure and memory company with three businesses: Advanced Computing (AI/HPC systems and services), Integrated Memory, and Optimized LED. Right now, the real growth engine is memory, while Advanced Computing is in the middle of a strategic reset away from hyperscaler concentration and toward enterprise, neocloud, and sovereign AI customers. That shift is hurting near-term reported growth because large deployments are timing-sensitive and Penguin Edge is being wound down, but management argues the quality of revenue should improve over time. The near-term story is straightforward: memory demand and pricing are lifting results and full-year guidance, while the call tries to convince investors that the weaker Advanced Computing outlook is a timing issue, not a demand issue, and that inference-focused AI products like CXL-based memory systems could become a higher-value growth layer on top of the legacy memory business.
Quarterly Results
Earnings Release Date: Apr. 1, 2026
Stock Price: $17.60
Market Cap: $931.0 million
Q2 2026 sales of $343.0 million vs $365.5 million in the prior year
Q2 2026 Non-GAAP Adjusted EPS of $0.52 vs $0.52 in the prior year
Q2 2026 GAAP Diluted EPS of $0.58 vs $0.09 in the prior year
Quick Takeaway
Penguin Solutions is in a growth-and-transition phase, trying to shift from a more concentrated and lumpier AI hardware profile toward a broader AI infrastructure and memory platform story. The near-term positives are strong memory demand, raised full-year guidance, improving customer diversification, and early proof points around inference-oriented products. The main concerns are that much of the current uplift appears pricing-driven, Advanced Computing is still under pressure from timing and transition issues, and the most exciting parts of the story are still ahead of the numbers.
Press Release vs Call Transcript Comparison
The call does a lot more work than the press release to explain why Advanced Computing should not be judged solely on the reported quarterly decline. That matters because a 42% YoY drop in the segment is obviously weak by any normal benchmark, but management is effectively asking investors to ignore the headline and focus on the underlying non-hyperscale business, which grew 50% in the first half and doubled its mix contribution year over year. If investors buy that argument, the stock can be treated like a transition story rather than a deteriorating one.
At the same time, the call is more candid than the press release about near-term margin pressure. Full-year gross margin guidance was cut even as EPS guidance rose, which usually means mix and below-the-line help are offsetting weaker profitability. Here, that mix issue comes from faster growth in lower-margin memory and AI hardware, plus higher memory input costs. For investors, that is neutral to mildly negative in the near term: revenue quality is improving in one sense, but not yet in reported margin structure.
Another important contrast is that the press release tells you Penguin has a MemoryAI product, while the call tries to show why it could matter strategically. The company is not just pitching a single product win. It is pitching a broader thesis that inference AI needs more memory, more low-latency architecture, and more integrated deployment expertise, which could let Penguin move from lower-value modules toward higher-value systems, software, and services over time. That thesis is promising, but still early. Investors need proof in future revenue mix, gross margin, and repeat customer wins.
Investor Underappreciation Signals
✅Advanced Computing Is Healthier Than the Headline — Reported Advanced Computing revenue looks weak, but the call shows the non-hyperscale AI/HPC business grew 50% in the first half and now makes up more than 40% of the segment, suggesting the market may be overreacting to a transition away from hyperscalers and Penguin Edge.
✅CXL Could Lift Margins, Not Just Revenue — The press release highlights the Tier One bank deployment, but the call adds that CXL systems should be higher margin than the traditional module business, which means investors may be underestimating the earnings power if adoption scales.
✅Bookings May Be Slipping, Not Breaking — Management said lower Advanced Computing guidance is mainly a timing issue tied to 3-to-6-month revenue lag and longer deployment cycles, so investors may be pricing in demand weakness that could reverse once projects convert.
✅Enterprise AI Diversification Is Becoming Real — The release mentions five new customers, but the call shows wins across financial services, biomedical research, energy, healthcare, retail, enterprise, neocloud, and sovereign AI, which could change perception from “customer concentration story” to “broadening platform story.”
✅Balance Sheet Flexibility Is a Competitive Asset — Investors may focus on segment noise and miss that Penguin ended in a net cash position, used proceeds from Celestial AI, retired debt, repurchased shares, and is using the balance sheet to secure constrained memory supply, which can help it out-execute smaller peers.
✅Inference May Become the Valuation Bridge — The press release hints at inference, but the call makes it central to the thesis, and if investors start viewing Penguin as a beneficiary of the AI shift from training to inference, the stock could get a narrative boost before the numbers fully catch up.
Section 2: Supplementary Information
Positive Insights
Negative Insights
Tariff Risk
Tariffs were not a major strategic discussion point on the call, but they were mentioned in an economically relevant way. The CFO said Q2 gross margin benefited in part from “tariff recovery in LED,” and that second-half gross margins are expected to be lower partly because of “less tariff cost recovery in LED.” That implies tariffs are affecting profitability in that segment, but the company currently has at least partial ability to recover some of those costs through pricing or pass-through mechanisms.
What the transcript does not show:
Management did not discuss broad tariff-driven supply-chain relocation, production shifts, renegotiated contracts, or any major effect on market share or product innovation. They also did not frame tariffs as the main risk to the business overall. The bigger operational pressure discussed on the call was memory cost inflation and supply constraints, not tariffs.
Hot Stock Trends Analysis
Previous Earnings Call
Quarter-over-quarter comparison (Previous Analysis)
In Q1 2026, Penguin Solutions framed itself as a company navigating a transition away from hyperscaler dependence while building a broader, more diversified customer base. Management emphasized pipeline growth, early enterprise adoption, and second-half revenue conversion, asking investors to focus on underlying momentum despite near-term variability.By Q2 2026, the narrative shifted toward positioning Penguin as an “AI factory platform” company aligned with the industry’s move from training to inference workloads. Management highlighted stronger customer validation, new inference-focused products, and raised full-year guidance, while also acknowledging timing-related weakness in Advanced Computing and a heavier reliance on memory pricing to drive near-term results.
Year-over-year comparison
In Q2 2025, Penguin Solutions presented itself as a high-growth AI infrastructure provider benefiting from hyperscale demand and early enterprise adoption, with strong revenue growth, margin expansion, and clear operating leverage. The narrative was straightforward: execution was strong, large deployments were driving results, and the company was scaling alongside the broader AI buildout cycle.
By Q2 2026, the story shifted to a transition phase, where Penguin is repositioning itself as an AI factory platform focused on inference and memory-centric architecture rather than hyperscale-driven growth. While management highlighted stronger enterprise traction and new products like CXL-based memory systems, the narrative now includes more complexity, with declining Advanced Computing revenue, reliance on memory pricing, and a greater emphasis on long-term strategic positioning over near-term financial consistency.