I stumbled across this company on X and was curious what they do. What if I told you this $3B company is quietly building national-scale AI infrastructure and that most retail investors have never heard of it? I’m currently long and wanted to give you an outline of what they do, bull case , bear case and let you decide.
Penguin Solutions is a “picks and shovels” play on AI. They don’t build AI models or make chips, they build and run the AI data centers that everyone else depends on. They run racks of servers, the chips inside, networking, cooling, power and the software tying it together. They build and run the factory: they don’t make what it produces. Penguin Solutions has a track record with 85,000+ GPUs deployed over their history.
The main two parts of their business are advanced computing and integrated memory. The advanced computing consists of designing and building the AI clusters. The integrated memory consists of specialized memory modules which are fast-access data storage for servers. This segment grew 63% last quarter due to AI demand and favorable pricing.
Here is what makes the story interesting. SK Telecom, a major Korean telecom giant, made a $200M strategic investment last year. Also, a collaboration with SKT + SK hynix to build AI data centers in Asia. They’re a core partner on “Haein”, one of Korea’s largest GPU-as-a-Service clusters. This is part of Korea’s national sovereign AI program.
Q2 revenue was 343M, EPS $0.52 vs $0.37 expected (40% beat), and raised full-year guidance (12% growth). We have to acknowledge that Q1 was a miss and the results could be lumpy going forward. The memory segment was +63% on AI demand and pricing.
I am going to now present the bull case. Every country wants its own AI infrastructure. PENG has a proven seat at the table via SKT/Haein. SKT‘s investment is validation and a pipeline into Asia-Pacific and Middle East. The 85,000+ GPUs and 2B+ runtime hours shows that they can actually deliver at scale. The memory segment is currently ripping straight into AI demand. They recently raised guidance which signals management confidence. If PENG becomes a go-to builder for national/enterprise AI factories globally, current revenue is a fraction of the opportunity- and that’s not priced in if it plays out.
Here comes the bear case. The stock is not cheap and is near the highs after a massive run. The average analyst target is $50 which is below the current price. Stifel actually has a $24 target which shows how divided the Street is over what this is worth. PENG’s current trailing PE is 90x. The stock has a high beta so be prepared for a bumpy ride in both directions. PENG’s CFO recently resigned and leadership transition is never easy. Short interest is currently 17% of the float which shows bearish positioning. Also, the next earnings are July first which is a near term catalyst and risk.
I’m currently bullish because I think the AI-infrastructure thesis outweighs the valuation risk. I think that it is priced for growth even though the Street is divided. Growth stalling and the SKT pipeline failing to expand would definitely change my mind. It’s not a huge market cap company so I’m aware of the downside risks. I just wanted to highlight this $3B company that hasn’t been talked about much and why I think it can keep climbing.
Disclosure: Long PENG. Not financial advice- do your own work.