Tesla (TSLA.US) is accelerating the development of its self-designed chips! The design of the AI5 chip is nearly complete, and the development of the AI6 chip has also commenced.

Tesla CEO Elon Musk said the company has largely completed design of its fifth-generation autonomous driving AI chip (AI5) and has begun early-stage development of AI6, targeting a nine-month chip design cycle. AI5 is slated for mass production in 2027 with Taiwan Semiconductor Manufacturing Co., and Tesla earlier signed a $16.5 billion deal with Samsung to produce A16 chips in the U.S.; the moves are intended to accelerate Tesla’s in-house silicon roadmap and reduce reliance on NVIDIA while signaling a recruitment push to scale production.

Analysis

Market structure: Tesla (TSLA) accelerating in-house AI chip design (AI5→AI6, nine‑month cadence) favors Tesla, its chosen foundry partners (TSM, SSNLF), and specialized automotive inference vendors while pressuring incremental automotive GPU TAM for NVIDIA (NVDA) by a low‑to‑mid single‑digit percent over 3–5 years. Expect upward pressure on foundry capacity and component lead times into 2026–2028 (Samsung $16.5bn deal + TSMC engagement), with equity vol rising in semis and modest tightening of TSLA credit spreads if execution looks credible. FX and commodity moves will be second‑order; stronger USD funds fab capex and Taiwan/SK flows, while copper/lithium impacts are minimal near term.

Risk assessment: Tail risks include catastrophic yield failures at TSMC/Samsung, a missed AI5 performance target, US export/restrictions or antitrust scrutiny of Tesla vertical integration, and a capital‑intensive pivot that compresses TSLA FCF. Immediate (days) risk is headline‑driven volatility; short term (weeks–months) risk is hiring/manufacturing noise; long term (2027+) execution and software ecosystem maturity determine value capture. Hidden dependency: Tesla must scale data center training and software tooling (not just inference silicon), and delays here nullify chip upside.

Trade implications: Tactical: establish a 2–3% long TSLA equity position now to capture upside from credible chip progress and sell near‑term covered calls to finance carry; add asymmetric exposure with 18–30 month LEAP calls (Jan 2028/29) sized ~1% net portfolio. Hedge: buy a 3–6 month NVDA 10–20% OTM put spread sized at ~50–70% of TSLA delta to protect against a headline‑driven drawdown. Play supply chain: initiate a 2–3% position in TSM (TSM) or SSNLF for 12–36 months to capture foundry scarcity; tighten stops to 12–15%.

Contrarian angles: The market may overstate NVDA displacement — NVIDIA’s datacenter moat (training GPUs, software stack) is durable and Tesla silicon is likely optimized for in‑car inference, not cloud training, so NVDA downside is limited. Conversely, consensus may underprice execution risk and capex strain on Tesla: Apple‑style vertical wins often take multiple product cycles and capex commitments. Unintended consequence: accelerated in‑house chips could divert engineering focus from vehicle product cadence and worsen short/medium‑term margins.

Key Decisions for Investors

• Establish a 2–3% long position in TSLA common stock within 1–4 weeks to capture optionality from AI5 progress; add ~1% via Jan 2028 LEAP calls (18–30 month expiry) as asymmetric upside exposure, and set a mental stop at -25% from entry on the equity leg.
• Hedge downside to that TSLA exposure by buying a 3–6 month NVDA 10–20% OTM put spread sized at 50–70% of the TSLA position delta (limit cost to <0.5% portfolio), protecting against headline‑driven sector repricing.
• Initiate a 2–3% position in TSM (TSM) or SSNLF (Samsung ADR) as a 12–36 month play on foundry benefit from Tesla’s chip program; target +20–30% if capacity tightens and trim/lock gains at +30% or cut at -15%.
• If overweight NVDA, trim 5–10% of NVDA exposure over the next 2–6 weeks to de‑risk headline volatility; redeploy proceeds into the TSLA/TSM exposures above, keeping net tech beta roughly neutral.