Samsung’s Q1 2026 edge over TSMC was an AI memory win, not a foundry flip

The three terms that prevent confusion

• DS, or Device Solutions: This is Samsung’s broad semiconductor reporting segment. In Q1 2026, Samsung listed DS Semiconductor sales of KRW 81.7 trillion, up 86% quarter over quarter, with operating profit of KRW 53.7 trillion and an operating margin of 66% .
• HBM and AI memory: HBM means high-bandwidth memory, a type of memory used in AI data centers. Samsung said DS record earnings were led by the memory business and increased sales of high-value-added AI products . External reporting also linked Samsung’s strong quarter to HBM sales for AI data centers .
• Foundry: This is the business of manufacturing chips for outside customers. Chosun described TSMC as the world’s largest foundry, meaning contract chip manufacturing company . SemiWiki still describes TSMC as the leader in pure-play foundry manufacturing, citing its manufacturing focus, customer trust, advanced nodes, yields and advanced packaging execution .

The headline numbers are answering different questions

One accounting detail matters: Samsung DS’s 66% figure is an operating profit margin, while TSMC’s 66.2% figure is a gross margin. TSMC’s operating margin for the same period was 58.1% . Those are not the same metric, so they should not be used as a direct scoreboard for foundry competitiveness.

Why Samsung looked ahead in this quarter

1. Memory did the heavy lifting

Samsung’s official materials say DS record earnings were led by the memory business, with higher sales of high-value-added AI products . Samsung’s news release also said the Memory Business set all-time highs for quarterly revenue and operating profit . That points to a product-cycle win in memory, not a foundry-only win.

2. AI demand lifted the value of memory chips

AI hardware is not only about advanced logic processors. It also needs large amounts of high-performance memory. External reporting tied Samsung’s Q1 2026 profit surge to strong sales of HBM chips used in AI data centers . That aligns with Samsung’s own explanation that high-value-added AI products helped drive DS performance .

3. Samsung’s profit was heavily concentrated in DS

Samsung DS reported KRW 53.7 trillion in operating profit, compared with Samsung Electronics’ total operating profit of KRW 57.2 trillion . Chosun reported that DS accounted for about 94% of Samsung Electronics’ total operating profit in Q1 2026, with semiconductors, especially memory, doing most of the work .

Why this is not proof that Samsung Foundry caught TSMC

First, the subject of the comparison is Samsung DS, not Samsung Foundry alone. Chosun’s claim was that Samsung’s DS semiconductor division appeared to surpass TSMC in overall scale and profitability during the memory super-boom . Samsung’s own materials emphasize memory and high-value AI products as the DS earnings driver .

Second, the foundry business did not show the same momentum in available reporting. Gadgets360 reported that Samsung’s foundry earnings declined in the first quarter, which sits awkwardly beside the broader DS record . In other words, DS can have a blowout quarter while Samsung Foundry remains a separate competitive question.

Third, TSMC was not having a weak quarter. It reported US$35.90 billion in revenue, a 66.2% gross margin and a 58.1% operating margin, with advanced nodes making up most wafer revenue . Its Q1 revenue also came in slightly above its original guidance range .

Industry analysis still treats TSMC as the pure-play foundry leader because of its focus on manufacturing, its lack of direct chip-design competition with major fabless customers, and its execution in advanced nodes and advanced packaging . Those are the areas that matter most when judging foundry strength.

What would actually show a foundry catch-up?

To argue that Samsung Foundry has truly caught TSMC, investors and customers would need same-scope evidence: more external foundry customer wins, stable volume production at advanced nodes, yields and delivery schedules that large customers trust, advanced packaging that scales alongside leading-edge processes, and several quarters of meaningful foundry profitability.

Those are different tests from asking whether Samsung DS had a huge memory-driven quarter. The latter is clearly supported by Q1 2026 results . The former is not proven by this comparison alone .

Bottom line

Samsung’s Q1 2026 performance was a major financial win for its semiconductor division. But it was mainly an AI memory win: DS surged on memory and high-value-added AI products, while TSMC continued to show strong foundry economics and advanced-node revenue . The quarter shows how powerful the AI memory cycle has become. It does not, by itself, prove that Samsung Foundry has beaten TSMC in chip manufacturing .