This NVIDIA Partner Is Already Up 123% in 2026, but Agentic AI Could Drive Its Next Growth Wave
Penguin Solutions CEO Kash Shaikh is making a bold bet that the real chokepoint in AI factories has nothing to do with GPUs, and the company's financials suggest Wall Street is only beginning to price in what agentic AI could…
Penguin Solutions CEO Kash Shaikh appeared on CNBC on July 29 with his thesis for the current AI cycle: “Memory is the new compute, especially with agentic AI.” As autonomous AI agents evolve from short prompt-and-response interactions into workloads operating around the clock, he expects the primary bottleneck in an AI factory to increasingly shift toward memory bandwidth and capacity rather than GPU throughput alone.
Shaikh described his company plainly: “Penguin Solutions is an AI factory platform company. We sit at the intersection of two very high-demand markets, AI infrastructure and memory.” He added that “enterprises, governments around the world and the new cloud providers are racing to build the AI factories” and that backlogs now extend multiple quarters.
Revenue Soared 48% as AI and Memory Demand Exploded
Penguin Solutions (NASDAQ:PENG) has become one of the most direct public-market vehicles for the memory-as-bottleneck thesis. Shares are up 123% since the start of 2026, with the company supporting a market cap of nearly $2.47 billion and analysts carrying a Buy consensus with a $74.29 price target, implying meaningful upside from the stock’s current price of $43.70.
The fundamentals back the CEO’s confidence. In fiscal Q3 2026, company-wide revenue grew 48% year over year, and the memory and AI infrastructure business grew over 104% year over year to represent over 75% of total net sales. Q3 saw revenue of $478.71 million, and non-GAAP diluted EPS of $0.84, beating consensus by 13.61% and 49.33%, respectively.
Management responded by raising fiscal 2026 net sales growth guidance to 22% ±2% and non-GAAP EPS guidance to $2.60 ±$0.05. Penguin was also recently named an NVIDIA AI Factory Specialized Partner and Dell’s Global Alliances Americas AI Partner of the Year.
Why Agentic AI Could Make Memory the Next Great Bottleneck
CEO Shaikh’s argument turned to how agentic workloads behave. Where advisory AI answers a question and stops, agentic AI is “performing tasks, automating workflows, and it is working 24/7.” Continuous context windows, persistent KV caches, and long-running tool use all pile pressure onto memory subsystems. Penguin’s MemoryAI CXL-based KV cache server, already deployed at a Tier One financial institution, is designed for exactly that workload.
On the earnings call, Shaikh reinforced the point, noting that “as inference and agentic AI workloads become more persistent and context-rich, memory is increasingly becoming one of the primary performance and scalability bottlenecks.”
Micron’s Historic Growth Validates the Memory Supercycle
Micron Technology (NASDAQ:MU | MU Price Prediction) offers a readout of the same phenomenon. Fiscal Q3 2026 revenue reached $41.46 billion, up 345.7% year over year, with GAAP gross margin expanding to 84.6%. CEO Sanjay Mehrotra told investors the results “reflect the strategic value of memory in the AI era.” HBM4 is now in high-volume shipments, and Micron guided Q4 revenue to $50.0 billion ±$1.0 billion. Shares are up 187.67% year to date.
NVIDIA Remains the Engine Behind the AI Factory Buildout
NVIDIA (NASDAQ:NVDA) remains the demand engine behind AI factory buildouts, with fiscal Q1 2027 revenue of $81.62 billion and Data Center revenue of $75.25 billion. Jensen Huang has called it “the largest infrastructure expansion in human history.” Penguin sits directly inside that ecosystem as an NVIDIA AI Factory Specialized Partner, and the two companies’ networking and memory roadmaps are increasingly coupled.
Penguin’s Biggest Risk Is Also Its Biggest Opportunity
Penguin trades at a forward P/E near 12, but the stock’s beta of 2.83 and a recent 22.68% one-month drawdown make it clear that investors are weighing memory-pricing risk against secular demand. While 74% of revenue is AI-related, about 89% of operating profit comes from the memory segment, meaning that Penguin is tied to the same cycle Micron rides.
If Shaikh is right that agentic AI will make memory a primary infrastructure bottleneck, that concentration could become Penguin’s greatest advantage. The next signals to watch are how quickly its multi-quarter backlog converts into revenue and whether MemoryAI CXL deployments expand beyond the initial Tier One customer.
Contact [email protected] for any questions or corrections.
Thomas Richmond is a financial writer and content strategist with 5+ years of experience covering stocks and financial markets. He has published over 500 articles focused on individual stock analysis, helping investors better understand business fundamentals, stock valuations, and long-term opportunities.
Thomas previously served as a Content Lead at TIKR, a stock research platform, where he helped scale the company’s blog to hundreds of articles per month and contributed to a weekly newsletter reaching more than 100,000 investors.
Outside of work, Thomas enjoys weight lifting and soccer.
