Terafab Project Accelerates Implementation, but at What Cost to Musk? Tech Analyst Ming-Chi Kuo: Terafab Faces Multiple Execution Risks

TradingKey - News has emerged that Elon Musk's self-developed chip project, Terafab, is accelerating its implementation. Photolithography equipment manufacturer ASML CEO Christophe Fouquet stated this week that he has held direct talks with Musk regarding the project, noting that Musk is "very serious" about it.

The latest industry survey released by renowned Hong Kong analyst Ming-Chi Kuo shows that this unprecedented project faces pressure in at least three areas: excessive project scope, an extremely tight timeline, and insufficient human resources. This will lead to consequences such as premiums on equipment procurement, the risk of missing technology R&D windows, and increased coordination difficulties from pursuing multiple process routes simultaneously. Given these overlapping pressures, Kuo believes that actual execution will be the deciding factor in the project's success.

Kuo believes that because Terafab is currently under immense time pressure, it needs to trade money for time. In his analysis, he noted that Terafab has already offered equipment manufacturers bids significantly higher than market rates for critical machinery. However, this could lead to a substantial increase in project costs, squeezing budgets in other areas.

Due to the tight schedule, Terafab's chip design window is extremely narrow. In terms of manufacturing processes, Terafab is utilizing Intel's (INTC) 14A node. It is expected that Intel's PDK 0.9 version will not be released to external customers until October 2026, which is when Terafab can begin actual design work for 14A. If Terafab fails to master it immediately, it could miss the small-batch production phase in 2028, fail to meet the 2028 pilot production target, or even fall behind by an entire process generation.

On the human resources front, Terafab faces equally severe pressure. According to Kuo's estimates, Apple's (AAPL) Silicon Engineering Group (SEG) is larger than the combined SpaceX and Tesla (TSLA) related chip design teams by dozens of times, yet the latter not only have shorter engineering timelines but also cover a broader range of chip development tasks.

Furthermore, regarding the scope of the project itself, Terafab represents an unprecedented challenge for a tech company. In terms of manufacturing collaboration, Terafab needs to pursue three advanced process paths simultaneously with TSMC (TSM) , Samsung, and Intel, a first for the IC design industry. Although Musk has long established a partnership with Intel, Terafab still needs multiple backup plans given Intel's yield issues. Additionally, SpaceX and Tesla have different requirements for their chips, necessitating matches with different foundries. This has resulted in a staggering workload for Terafab.

At the product line level, the project needs to cover two major computing power lines: those for ground-based use (edge inference) and those optimized specifically for space environments, as well as more than six parallel chip projects including the AI series, Dojo series, and SpaceX-exclusive chips.

Regarding the chip R&D and manufacturing process, Terafab plans to integrate four key processes—mask design, logic, memory chips, and advanced packaging—within a single facility. This level of vertical integration is virtually unprecedented in the current industry. The semiconductor industry typically employs a regional division of labor for these four processes—for example, the U.S. handles design, Taiwan fabs the logic, South Korea manufactures memory, and Malaysia performs packaging—balancing cost and efficiency. Musk's extreme vertical design maximizes efficiency and engineering quality but could significantly drive up factory construction costs.

In terms of the chip design cycle, Terafab aims to complete a design iteration in 9 months, whereas the industry norm is 18 to 24 months, with complex designs often requiring 2 to 3 years. This means Terafab's iteration speed must be more than double the industry average.

Kuo believes that under these circumstances, Musk needs to find the best partners to mitigate risks at every stage, and MediaTek would be an excellent choice. MediaTek has practical experience collaborating with Intel, including a track record of tape-outs for the Intel 16 process and involvement with EMIB-T advanced packaging, which could help Terafab quickly ramp up chip design. MediaTek has also long supplied Wi-Fi/Router SoCs for Starlink user terminals, establishing a relationship of commercial trust with Musk's SpaceX. Nevertheless, the more critical factor remains Musk's actual execution capability: how he will push progress and solve problems at crucial moments in the face of costs and engineering speeds far exceeding industry averages.