News
SpaceX’s BFR factory abuzz with work activity and giant rocket tooling
A large temporary SpaceX facility intended for initial Mars rocket fabrication was spied abuzz with activity as employees work to assemble massive carbon composite tooling.
Hours after SpaceX CEO Elon Musk revealed the first photos of tooling for the manufacture of BFR, Teslarati photographer Pauline Acalin sped down to Port of San Pedro to survey a large dockside tent the company was constructing as of December 2017. Now complete, the temporary facility appears to be exploding with SpaceX activity as the company surges ahead with plans to assemble the first Mars rocket and spaceship prototypes – rocket-powered hops could begin as soon as early next year, with orbital launches following about a year after that.
@SpaceX showing off the main body tool for the BFR 🤩🤩🤩https://t.co/IvvN7z9kx5 pic.twitter.com/GqXssPJhqG
— Ascent Aerospace (@AscentAerospace) April 9, 2018
Per a number of related discoveries, the tooling pictured in Musk’s teaser is almost certainly located in the same tent pictured above. Of particular note, a source involved in the work has confirmed that SpaceX is using a new supplier for the custom tooling needed to manufacture BFR. The source’s comments were confirmed to be accurate minutes later in photos taken by Teslarati’s senior SpaceX stalker that peg Ascent Aerospace Coast Composites as the tooling manufacturer. As if to dispatch any lingering doubt, Ascent Aerospace appears to have also independently confirmed its involvement through a rare post on social media.
Based on Pauline Acalin’s photos of the previously unexplored SpaceX facility, the company has since filled the tent with a huge amount of hardware, and still more BFR tooling appears to be momentarily stored outside, brandishing the Ascent Aerospace name and logo. These additional components will likely be used in the manufacture of the less cylindrical segments of SpaceX’s Mars spaceship, among other possible uses. In essence, SpaceX will need to build monolithic carbon composite structures that can readily survive extreme temperatures, pressure differentials, supercool rocket propellant, significant G-loading, and much more. To a much lesser extent, this type of aerospace construction is already done on a large scale – Boeing’s 787 Dreamliner uses a significant amount of carbon composite in its fuselage and is around 6 meters in diameter to BFR/BFS’ 9m. As a beneficial consequence of such significant commercial interest in massive carbon composite structures, markets have grown to support the existence of a thriving niche of composite innovation and tooling production, maturing the technology and making the task far easier for SpaceX compared to developing tooling and processes from a blank sheet.
- SpaceX appears to be hard at work at a temporary tent believed to house initial BFR manufacturing facilities. (Pauline Acalin)
- SpaceX’s first major BFR and BFS fabrication tooling, likely being stored temporarily in a tent at Port of San Pedro. Note the tent framework at the top. (Elon Musk)
- It’s understood that SpaceX will eventually move this work to Berth 240 once more permanent facilities are constructed. (Pauline Acalin)
- Huge metal structures are being stored just outside a tent constructed for SpaceX’s initial BFR manufacturing. (Pauline Acalin)
Given the surprising level of activity at this BFR-focused facility, it is fair to conclude that SpaceX is wasting no time at all with the production of its first full-scale BFR prototypes. Altogether, the tent factory is gradually being filled to the brim with custom carbon composite tooling capable of fabricating Mars spaceship and booster structures, propellant tanks, delta winglets, and more. It’s no coincidence that this tent (and the prospective factory at Berth 240) were both located at Port of San Pedro – once completed, it should be comparatively easy to ship the massive components to SpaceX’s Texas or Florida facilities, both of which have been hinted as possible locations for BFR testing (and launches, eventually).
Follow us for live updates, behind-the-scenes sneak peeks, and a sea of beautiful photos from our East and West coast photographers.
Teslarati – Instagram – Twitter
Tom Cross – Twitter
Pauline Acalin – Twitter
Eric Ralph – Twitter
News
Tesla shares AI5 chip’s ambitious production roadmap details
Tesla CEO Elon Musk has revealed new details about the company’s next-generation AI5 chip, describing it as “an amazing design.”
Tesla CEO Elon Musk has revealed new details about the company’s next-generation AI5 chip, describing it as “an amazing design” that could outperform its predecessor by a notable margin. Speaking during Tesla’s Q3 2025 earnings call, Musk outlined how the chip will be manufactured in partnership with both Samsung and TSMC, with production based entirely in the United States.
What makes AI5 special
According to Musk, the AI5 represents a complete evolution of Tesla’s in-house AI hardware, building on lessons learned from the AI4 system currently used in its vehicles and data centers. “By some metrics, the AI5 chip will be 40x better than the AI4 chip, not 40%, 40x,” Musk said during the Q3 2025 earnings call. He credited Tesla’s unique vertical integration for the breakthrough, noting that the company designs both the software and hardware stack for its self-driving systems.
To streamline the new chip, Tesla eliminated several traditional components, including the legacy GPU and image signal processor, since the AI5 architecture already incorporates those capabilities. Musk explained that these deletions allow the chip to fit within a half-reticle design, improving efficiency and power management.
“This is a beautiful chip,” Musk said. “I’ve poured so much life energy into this chip personally, and I’m confident this is going to be a winner.”
Tesla’s dual manufacturing strategy for AI5
Musk confirmed that both Samsung’s Texas facility and TSMC’s Arizona plant will fabricate AI5 chips, with each partner contributing to early production. “It makes sense to have both Samsung and TSMC focus on AI5,” the CEO said, adding that while Samsung has slightly more advanced equipment, both fabs will support Tesla’s U.S.-based production goals.
Tesla’s explicit objective, according to Musk, is to create an oversupply of AI5 chips. The surplus units could be used in Tesla’s vehicles, humanoid robots, or data centers, which already use a mix of AI4 and NVIDIA hardware for training. “We’re not about to replace NVIDIA,” Musk clarified. “But if we have too many AI5 chips, we can always put them in the data center.”
Musk emphasized that Tesla’s focus on designing for a single customer gives it a massive advantage in simplicity and optimization. “NVIDIA… (has to) satisfy a large range of requirements from many customers. Tesla only has to satisfy one customer, Tesla,” he said. This, Musk stressed, allows Tesla to delete unnecessary complexity and deliver what could be the best performance per watt and per dollar in the industry once AI5 production scales.
Energy
Tesla VP hints at Solar Roof comeback with Giga New York push
The comments hint at possible renewed life for the Solar Roof program, which has seen years of slow growth since its 2016 unveiling.
Tesla’s long-awaited and way underrated Solar Roof may finally be getting its moment. During the company’s Q3 2025 earnings call, Vice President of Energy Engineering Michael Snyder revealed that production of a new residential solar panel has started at Tesla’s Buffalo, New York facility, with shipments to customers beginning in the first quarter of 2026.
The comments hint at possible renewed life for the Solar Roof program, which has seen years of slow growth since its 2016 unveiling.
Tesla Energy’s strong demand
Responding to an investor question about Tesla’s energy backlog, Snyder said demand for Megapack and Powerwall continues to be “really strong” into next year. He also noted positive customer feedback for the company’s new Megablock product, which is expected to start shipping from Houston in 2026.
“We’re seeing remarkable growth in the demand for AI and data center applications as hyperscalers and utilities have seen the versatility of the Megapack product. It increases reliability and relieves grid constraints,” he said.
Snyder also highlighted a “surge in residential solar demand in the US,” attributing the spike to recent policy changes that incentivize home installations. Tesla expects this trend to continue into 2026, helped by the rollout of a new solar lease product that makes adoption more affordable for homeowners.
Possible Solar Roof revival?
Perhaps the most intriguing part of Snyder’s remarks, however, was Tesla’s move to begin production of its “residential solar panel” in Buffalo, New York. He described the new panels as having “industry-leading aesthetics” and shape performance, language Tesla has used to market its Solar Roof tiles in the past.
“We also began production of our Tesla residential solar panel in our Buffalo factory, and we will be shipping that to customers starting Q1. The panel has industry-leading aesthetics and shape performance and demonstrates our continued commitment to US manufacturing,” Snyder said during the Q3 2025 earnings call.
Snyder did not explicitly name the product, though his reference to aesthetics has fueled speculation that Tesla may finally be preparing a large-scale and serious rollout of its Solar Roof line.
Originally unveiled in 2016, the Solar Roof was intended to transform rooftops into clean energy generators without compromising on design. However, despite early enthusiasm, production and installation volumes have remained limited for years. In 2023, a report from Wood Mackenzie claimed that there were only 3,000 operational Solar Roof installations across the United States at the time, far below forecasts. In response, the official Tesla Energy account on X stated that the report was “incorrect by a large margin.”
News
Tesla VP explains why end-to-end AI is the future of self-driving
Using examples from real-world driving, he said Tesla’s AI can learn subtle value judgments, the VP noted.
Tesla’s VP of AI/Autopilot software, Ashok Elluswamy, has offered a rare inside look at how the company’s AI system learns to drive. After speaking at the International Conference on Computer Vision, Elluswamy shared details of Tesla’s “end-to-end” neural network in a post on social media platform X.
How Tesla’s end-to-end system differs from competitors
As per Elluswamy’s post, most other autonomous driving companies rely on modular, sensor-heavy systems that separate perception, planning, and control. In contrast, Tesla’s approach, the VP stated, links all of these together into one continuously trained neural network. “The gradients flow all the way from controls to sensor inputs, thus optimizing the entire network holistically,” he explained.
He noted that the benefit of this architecture is scalability and alignment with human-like reasoning. Using examples from real-world driving, he said Tesla’s AI can learn subtle value judgments, such as deciding whether to drive around a puddle or briefly enter an empty oncoming lane. “Self-driving cars are constantly subject to mini-trolley problems,” Elluswamy wrote. “By training on human data, the robots learn values that are aligned with what humans value.”
This system, Elluswamy stressed, allows the AI to interpret nuanced intent, such as whether animals on the road intend to cross or stay put. These nuances are quite difficult to code manually.
Tackling scale, interpretability, and simulation
Elluswamy acknowledged that the challenges are immense. Tesla’s AI processes billions of “input tokens” from multiple cameras, navigation maps, and kinematic data. To handle that scale, the company’s global fleet provides what he called a “Niagara Falls of data,” generating the equivalent of 500 years of driving every day. Sophisticated data pipelines then curate the most valuable training samples.
Tesla built tools to make its network interpretable and testable. The company’s Generative Gaussian Splatting method can reconstruct 3D scenes in milliseconds and model dynamic objects without complex setup. Apart from this, Tesla’s neural world simulator allows engineers to safely test new driving models in realistic virtual environments, generating high-resolution, causal responses in real time.
Elluswamy concluded that this same architecture will eventually extend to Optimus, Tesla’s humanoid robot. “The work done here will tremendously benefit all of humanity,” he said, calling Tesla “the best place to work on AI on the planet currently.”
-
Elon Musk1 week agoSpaceX posts Starship booster feat that’s so nutty, it doesn’t even look real
-
Elon Musk6 days agoTesla Full Self-Driving gets an offer to be insured for ‘almost free’
-
News6 days agoElon Musk confirms Tesla FSD V14.2 will see widespread rollout
-
News7 days agoTesla is adding an interesting feature to its centerscreen in a coming update
-
News1 week agoTesla launches new interior option for Model Y
-
News1 week agoTesla widens rollout of new Full Self-Driving suite to more owners
-
Elon Musk7 days agoTesla CEO Elon Musk’s $1 trillion pay package hits first adversity from proxy firm
-
News5 days agoTesla might be doing away with a long-included feature with its vehicles






