Elon Musk announces plans to manufacture chips for SpaceX and Tesla

Elon Musk announced a proposed 'Terafab' near Tesla HQ in Austin to build two dedicated chip fabs for Tesla/humanoid robots and for SpaceX/xAI space data centres, aiming to produce 1 terawatt of power a year. Musk said the fabs will make chips supporting 100–200 GW/year and others that can support a terawatt in space; no timelines were provided. The move is intended to address chip supply constraints for Musk’s projects and could materially affect Tesla/SpaceX supply chains and capex plans if pursued.

Analysis

This is a multi-year strategic bet, not a near-term earnings lever. Building vertically-integrated fabs shifts Tesla from a demand signal to a supply-creator — that increases total addressable demand for upstream silicon (substrates, packaging, specialty analog) while compressing margins for any incumbent supplier that relies on exclusivity of design wins. Expect a 24–36 month development horizon before production-scale output; during that window capital intensity, permitting, and equipment lead times (months to years) are the binding constraints.

Second-order winners are likely to be OSATs and power-IC specialists that sit between wafer fabs and final modules: if Tesla targets mature nodes with heavy power/packaging needs, packaging houses and power-IC vendors see concrete order flow before advanced EUV-dependent suppliers do. Conversely, advanced-node equipment (EUV) and datacenter GPU incumbents face limited immediate disruption because space-data-center and robotic AI workloads will be highly specialized and latency/thermal-constrained.

Regulatory and grid friction is an underpriced execution risk. A petawatt-scale local load in Texas invites interconnection delays, transmission upgrades, and potential political pushback; each can add 12–36 months and materially increase capex per watt. Finally, Musk’s move increases strategic optionality for Tesla: owning the stack reduces procurement risk but also puts Tesla in direct competition with suppliers for every incremental dollar of margin, changing its capital allocation calculus and potentially increasing cyclicality of FCF over a multi-year buildout.

Key Decisions for Investors

• Long TSLA LEAPS call spread (buy 2028 LEAP call / sell 2028 higher-strike call) to capture optionality from successful vertical integration over a 3–5 year horizon; limit premium risk to 100% of debit with upside skew 2–4x if Terafab materially improves gross margins and supply visibility.
• Long Amkor Technology (AMKR) or similar OSAT exposure for 12–36 months to play increased packaging demand from mature-node, high-power chips; target a 6–12% position size with a 30–40% stop, asymmetric upside if Tesla outsources initial volumes before in-sourcing.
• Long select power-IC specialists (e.g., ON Semiconductor - ON) on pullbacks for 6–24 months — these benefit if Tesla accelerates in-house power management designs but are also at risk of later in-sourcing; size as a tactical 3–5% position and use protective puts on >20% move against you.
• Tactical short (buy puts) on TSLA over 0–6 months around key permitting/announcement dates to hedge headline-driven rallies: execution and grid/permitting delays are high-probability catalysts that can compress multiple expansions; keep hedges modest (1–3% portfolio) as catalysts are binary and time-limited.