Samsung's Profit Nearly Octuples as AI Chip Demand Keeps Climbing
The South Korean giant also committed fresh money to AI data center infrastructure and struck a deal with French AI startup Mistral to help design its next generation of chips.

Samsung Electronics posted third-quarter results on Thursday that landed almost exactly where Wall Street and local analysts expected, with operating profit up 783% from a year earlier as the company keeps cashing in on the AI boom's appetite for advanced memory chips.
The company reported revenue of 195.25 trillion won, just shy of the 195.79 trillion won analysts had projected, and operating profit of 107.44 trillion won against expectations of 107.66 trillion won, according to LSEG SmartEstimates, a forecasting measure that weights more heavily toward analysts with stronger track records. Both figures came in close enough to consensus that the quarter reads as a confirmation of trend rather than a surprise.
The scale of the year-over-year swing is the real headline. Revenue climbed 126.88% from the same period last year and rose 13.85% from the prior quarter, while operating profit shot up 783.11% year-on-year. That kind of jump reflects how dramatically the memory chip market has turned in Samsung's favor since AI infrastructure spending took off. Companies building out AI systems need enormous amounts of high-performance memory to keep up with the data demands of training and running large models, and Samsung is one of a small number of manufacturers capable of supplying it at scale.
Betting bigger on AI infrastructure
Beyond the earnings print, Samsung used the quarter to signal it wants a bigger footprint in the infrastructure that AI runs on, not just the chips that power it. Samsung Electronics and five of its affiliated companies announced a combined $1 billion investment in Helix Digital Infrastructure, a data center platform backed by private equity firm KKR. Samsung Electronics itself is putting up $500 million of that total, with its affiliates covering the remainder.
Investing directly in data center infrastructure rather than just supplying components to it is a notable shift in posture. It gives Samsung a stake in the broader AI buildout beyond chip sales, tying its fortunes more directly to how much capacity companies like KKR's Helix end up deploying. For a company whose recent profit surge has been driven almost entirely by memory chip demand, it's a way to extend that exposure further up the value chain.
Mistral deal brings AI into chipmaking itself
Samsung also announced a partnership with Mistral AI, the French artificial intelligence startup, aimed at applying AI technology to Samsung's own semiconductor operations, including chip design and manufacturing. Samsung led Mistral's Series D funding round, putting the company on both sides of the relationship: an investor in Mistral's growth and a customer applying its technology internally.
The logic here is straightforward. Chip design and fabrication are enormously complex processes, and AI tools are increasingly being used across the industry to speed up design cycles and improve manufacturing precision. By partnering with Mistral rather than building everything in-house, Samsung gets access to a well-regarded AI model developer while also backing a startup that's become one of Europe's most prominent AI companies.
Taken together, the quarter tells a fairly clean story. The AI memory boom that has been reshaping Samsung's financials for several quarters is still running hot, with profit growth so steep it's nearly eightfold year-over-year. At the same time, Samsung is spending some of that windfall on positioning itself more broadly within the AI ecosystem: owning a piece of the data centers consuming its chips, and bringing AI tools into the factories that build them.
None of this guarantees the growth continues at this pace. Memory chip markets have historically been cyclical, and a 783% profit jump from a year ago is as much a reflection of how weak that prior period was as it is a sign of permanent altitude. But for now, Samsung's results confirm that the AI infrastructure buildout is still translating directly into its bottom line, and the company is moving to make sure it stays embedded in that buildout as it evolves.


