Tesla, Meta, and Anthropic: Samsung Electronics' chip foundry orders backlog has reached 50 trillion won.
```
Samsung Electronics is rapidly establishing its core position in the AI semiconductor market through its foundry business.
On July 3rd, according to Korean media reports, the backlog of long-term orders for Samsung’s foundry division has reached nearly 50 trillion KRW. After securing Tesla’s AI chip order last year, Meta and Anthropic, two global tech giants, have also shifted their ASIC production demand to Samsung.
An industry insider said:
Since Samsung’s foundry secured Tesla's AI chip order last year, semiconductor foundry orders for AI servers have entered a phase of full acceleration.
Driven by this, the market expects Samsung’s foundry business to turn profitable as early as the fourth quarter of this year. The key support for this round of order expansion lies in Samsung’s most advanced 2-nanometer process technology.
Meta’s third-generation AI accelerator chip MTIA 3 and Anthropic’s customized ASIC both plan to use this process for production, and external demand inquiries for Samsung’s 2nm process have surged, further solidifying its competitive position in the advanced process foundry market.
Meta shifts orders to Samsung, bets on 2nm mass production
According to reports citing industry insiders, Meta is negotiating with Samsung’s foundry division on next-generation ASIC design and production cooperation totaling over 10 trillion KRW.
The first two generations of Meta’s self-developed AI accelerator MTIA were manufactured by TSMC, but starting from the third generation released this year, Meta has designated Samsung as its core manufacturing partner.
According to reports, MTIA 3 will use Samsung’s most advanced 2nm process, with production scale reaching hundreds of thousands of wafers. Samsung Electronics stated that "nothing is decided yet".
Behind Meta’s shift to Samsung lies its strategic need for large-scale, self-built AI infrastructure.
Meta is researching cloud service businesses that lease AI computing power to external companies, and MTIA will serve as the core chip supporting this business.
Meanwhile, Meta has set a goal to build data centers with a total capacity of 5GW by 2030, meaning reliance solely on external chip supply becomes unsustainable.
For this, Meta has initiated a fast-paced development cycle, iterating new chips every six months, planning to release the third to fifth generations in succession by next year.
To meet this ultra-fast development cycle, Meta has also established a joint design mechanism with Samsung’s System LSI division. According to reports, both parties’ cooperation began from the early stages of chip architecture design, compensating for Meta’s engineering team’s capacity gaps under such a compressed schedule.
Anthropic internalizes AI infrastructure, Samsung may be the biggest beneficiary
US AI company Anthropic is reported to be evaluating the use of Samsung’s 2nm foundry process to develop custom ASICs, which is seen as part of its strategy to reduce reliance on Nvidia GPUs and Google TPUs, promoting "AI infrastructure independence".
In terms of investment scale, Anthropic's long-term plan is to build AI data centers with a capacity of about 1GW, with related investment estimated to be around 50 billion USD (about 77 trillion KRW).
Industry analysis believes that about half of these funds will go toward AI semiconductor procurement, with investments in ASICs, DRAM, and NAND flash estimated at around 25 billion USD (about 39 trillion KRW).
Samsung Electronics, possessing integrated capabilities in memory, foundry, and advanced packaging, is widely seen as the biggest potential beneficiary of Anthropic's current AI chip procurement.
In May this year, Samsung participated in Anthropic's $65 billion (about 100 trillion KRW) Series H financing, establishing a strategic partnership between the two sides.
Risk warning and disclaimerThe market carries risk and investment should be cautious. This article does not constitute personal investment advice, nor does it take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinion, viewpoint, or conclusion in this article fits their specific situation. Investing based on this information is at your own risk. ```