Отчётный звонок Tesla, Inc.
Unsupervised FSD in Austin within months, expanding to 8-10 metros by year-end
Tesla Q3 2025 call was anchored by Elon Musk's bullish narrative on unsupervised FSD launch and aggressive production expansion. Management highlighted AI5 chip design gains, tariff headwinds on energy storage, and Optimus as a future revenue juggernaut. Cross-company mentions centered on Samsung and TSMC for chip manufacturing, NVIDIA for training, and xAI/OpenAI/Google as AI competitors. Tesla expects to remove safety drivers in Austin by year-end and expand Robotaxi to 8-10 metros, signaling confidence in unsupervised FSD.
Вывод Buzzberg Unsupervised FSD in Austin within months, expanding to 8-10 metros by year-end Tesla Q3 2025 call was anchored by Elon Musk's bullish narrative on unsupervised FSD launch and aggressive production expansion. Management highlighted AI5 chip design gains, tariff headwinds on energy storage, and Optimus as a future revenue juggernaut. Cross-company mentions centered on Samsung and TSMC for chip manufacturing, NVIDIA for training, and xAI/OpenAI/Google as AI competitors. Tesla expects to remove safety drivers in Austin by year-end and expand Robotaxi to 8-10 metros, signaling confidence in unsupervised FSD. Читать полный анализСвернуть анализ
Tesla Q3 2025 call was anchored by Elon Musk's bullish narrative on unsupervised FSD launch and aggressive production expansion. Management highlighted AI5 chip design gains, tariff headwinds on energy storage, and Optimus as a future revenue juggernaut. Cross-company mentions centered on Samsung and TSMC for chip manufacturing, NVIDIA for training, and xAI/OpenAI/Google as AI competitors. Tesla expects to remove safety drivers in Austin by year-end and expand Robotaxi to 8-10 metros, signaling confidence in unsupervised FSD.
- AI5 chip design deletes legacy GPU/ISP, enabling half-reticle die and 40x improvement over AI4; both TSMC and Samsung will manufacture AI5 in US fabs.
- Tariff impact exceeded $400M in Q3, split evenly between auto and energy; Shanghai Megafactory helps mitigate for non-US markets.
- CapEx guided to ~$9B for FY2025, with substantial increases in 2026 for AI and Optimus production ramp.
Что важно сейчас
Самые значимые изменения по итогам звонка.
AI5 chip claims 10x performance per dollar, 40x better than AI4
Production expansion to 3 million annualized within 24 months
Показать ещё 3 тезиса
Optimus V3 to be production-intent prototype by Q1, million-unit line planned
Capex to increase substantially in 2026 for AI and Optimus
Megapack 4 to output 35 kV, reducing substation dependency
Фактические результаты
| Показатель | Факт | Изменение |
|---|---|---|
| Выручка | $28.095B | Факт |
| EPS | $0.50 | Факт |
| AUTOMOTIVE Валовая маржа | 15.4% | Факт |
| Валовая маржа | 17.99% | Факт |
| Операционная маржа | 5.78% | Факт |
| Свободный денежный поток | $3.99B | Факт |
Прогноз компании
| Показатель | Период | Диапазон | Середина | Статус |
|---|---|---|---|---|
| Капзатраты | FY2025 | $9B | $9B | Обозначен |
Оценка менеджмента
Bullish
Management expressed high confidence in achieving unsupervised FSD, expanding production rapidly, and highlighted transformative potential of Optimus and AI5 chip.
Оценка AI от менеджмента
Management emphasized Tesla's leadership in real-world AI, with FSD V14 released to US customers and a roadmap to add reasoning and increase parameter count by an order of magnitude. They highlighted the AI5 chip design, which they claim will be 40 times better than AI4 by some metrics, and noted the potential for using idle fleet compute as distributed inference. They also discussed plans to expa
Инвестиции и мощности
2025 capex is expected to be around $9 billion, with projections to increase substantially in 2026 to fund AI initiatives including Optimus and expansion of vehicle production to an annualized rate of 3 million within 24 months. They are also planning a million-unit Optimus production line, with production start towards the end of 2026.
Компаниис момента звонка
Поставщики
Samsung is confirmed as a key supplier for Tesla's AI4 computer and will also manufacture AI5 chips, reinforcing their relationship and potential revenue stream.
Доказательства
“I have nothing but great things to say about Samsung. They're an amazing company. And Samsung, it is worth noting, does manufacture our AI4 computer and does a great job doing that.”
Tesla continues to use NVIDIA GPUs for training, indicating sustained demand from Tesla for NVIDIA hardware despite developing proprietary chips.
Доказательства
“We already use AI4 for training in our data center. So we use a combination of AI4 and NVIDIA hardware. So we're not about to replace NVIDIA, to be clear, but we do use both in combination.”
Конкуренты
Tesla's CEO positions xAI's Grok as a direct competitor to OpenAI's ChatGPT, but no direct Tesla business impact.
Доказательства
“XAI and Grok are competing with Google Gemini and OpenAI, ChatGPT, and that kind of thing.”
Альфа цепочки поставок · 3с момента звонка
Tesla's AI5 chip design eliminates legacy GPU and ISP, enabling half-reticle size and simpler manufacturing, potentially improving yields and reducing cost vs. NVIDIA's complex designs.
Доказательства
“we deleted the legacy GPU or the traditional GPU, which is in AI4. ... we can actually fit AI5 in a half reticle with good margin for the traces from the memory to the ... Tesla trip accelerators.”
Tesla's energy storage business faces $400M+ quarterly tariff headwind, with all cells currently sourced from China, but the Shanghai Megafactory helps dodge tariffs for non-US markets.
Доказательства
“The total tariff impacts for Q3 for both businesses was in excess of 400 million, generally split evenly between them.”
Tesla plans to ramp Optimus production to a million-unit annualized rate by end of next year, requiring extreme vertical integration due to nonexistent supply chain.
Доказательства
“there is no supply chain. So in order to manufacture that, Tesla actually has to be very vertically integrated and manufacture very deep into the supply chain, manufacture the parts internally.”
Методология и полнота
Анализ основан только на заявлениях менеджмента. Показаны все подтверждённые упоминания компаний: 4. Фактические результаты и прогнозы разделены. Публичные доказательства ограничены восемью короткими атрибутированными цитатами. AI-анализ может быть неполным или ошибочным — проверяйте важные утверждения по первоисточнику.