Connect with us

News

Tesla’s Elon Musk shares first look inside Cortex supercluster at Giga Texas

Credit: Elon Musk/X

Published

on

Elon Musk recently shared a first look inside Cortex, Tesla’s supercomputer cluster at the Giga Texas complex. The short video, which shows a section of Cortex that was still being set up, was posted by the CEO on social media platform X. 

As could be seen in the video, the Giga Texas-based supercluster spans rows upon rows of Nvidia H100/H200 GPUs. Musk has previously noted on X that Cortex will feature about 100,000 H100/H200 GPUs, which should provide some serious muscle for Tesla’s AI training. “Video of the inside of Cortex today, the giant new AI training supercluster being built at Tesla HQ in Austin to solve real-world AI,” Musk wrote in his post on X.

Tesla’s artificial intelligence efforts are critical in the development of two of the company’s most important products, Full Self-Driving (FSD) and Optimus, a humanoid robot. The feasibility of Tesla’s planned Robotaxi network, which was outlined by Elon Musk in Master Plan, Part Deux way back in 2016, is highly dependent on the success of FSD. Optimus, meanwhile, has the potential to change numerous industries, such as manufacturing. 

Cortex is a very large supercomputer cluster, so much so that the facility required a whole new extension to Giga Texas’ main building. Musk also previously mentioned that Cortex will require about 130 MW of cooling this year before expanding to over 500 MW of cooling in the next 18 months. Drone flyovers of the Giga Texas facility show that work is underway to build a large supercomputer cooling system near Cortex. 

The Giga Texas-based supercluster is not the only supercomputer that will be built by Tesla for its FSD efforts. Earlier this month, Tesla signed an agreement with Buffalo, New York, to extend the company’s commitment to operate Giga New York for five more years until 2034. As part of the new agreement, Tesla will be investing $500 million in a Dojo supercomputer project in Buffalo, with $350 million being invested in the project by the end of next year. Dojo, as per Tesla’s previous presentations, is expected to help in the development of autonomous driving systems. 

Don’t hesitate to contact us with news tips. Just send a message to simon@teslarati.com to give us a heads up.

Advertisement
-

Simon is an experienced automotive reporter with a passion for electric cars and clean energy. Fascinated by the world envisioned by Elon Musk, he hopes to make it to Mars (at least as a tourist) someday. For stories or tips--or even to just say a simple hello--send a message to his email, simon@teslarati.com or his handle on X, @ResidentSponge.

Advertisement
Comments

Elon Musk

Elon Musk shuts down talk of TSMC taking over Terafab

Musk says Tesla and SpaceX will build and run Terafab, with TSMC limited to renting.

Published

on

By

SpaceX Terafab rendering

Elon Musk has drawn a firm line around who will be in charge of Terafab, the giant chip factory Tesla and SpaceX are planning in Texas.

Musk replied to a post on X arguing that Taiwan Semiconductor Manufacturing Company (TSMC) would most likely end up owning and operating the plant. “No, we will build and run the fab. Let there be ZERO doubt about that,” Musk wrote. “Maybe TSMC subleases part of the Terafab if they want, but nothing more than that.”

In plain terms, a sublease means TSMC could rent a section of the complex to make chips, similar to a tenant renting one floor of an office tower. The building, the equipment decisions and the daily operation would stay with Tesla and SpaceX.

The comment shuts down speculation that started last week. On October 2, tech journalist Tim Culpan reported that TSMC was exploring ways to help run Terafab’s factories. Musk responded the next day that it was “just discussions, but something may come of it,” as Teslarati reported at the time. That left room for a scenario where the world’s largest contract chipmaker took the wheel. Musk’s latest post closes that door.

Elon Musk teases TSMC as potential Terafab partner

Some background helps explain why this matters. Tesla designs its own AI chips today but pays outside companies like TSMC and Samsung to manufacture them. Musk unveiled Terafab in March as a joint project between Tesla, SpaceX and xAI, arguing that existing suppliers cannot expand fast enough to meet his companies’ future demand. The goal is to produce enough chips each year to supply one terawatt of computing power, roughly 50 times what the entire global AI chip industry produces now.

Those chips are meant for Tesla’s Optimus humanoid robots, the Cybercab and Full Self-Driving computers, along with chips for SpaceX’s planned data centers in orbit. Owning the factory means Musk’s companies would not have to compete with every other chip customer for time on someone else’s production lines.

Intel is still part of the picture. The company signed on in April to help design, build and package chips for the project, and CEO Lip-Bu Tan told Bloomberg this week that Intel will keep working on Terafab despite the TSMC chatter.

The project moved from concept to construction planning over the summer. In August, SpaceX confirmed the Grimes County site about an hour from Houston, sent the county a $10 million payment under its tax abatement deal and said civil work would begin shortly. The first phase carries a $16.8 billion price tag, and total spending across all phases could reach as much as $119 billion.

TSMC chairman C.C. Wei has said a new fab typically takes two to three years to build and another one to two years to reach full output. Tesla and SpaceX have never run one, which is why TSMC’s expertise drew so much attention. Musk’s answer suggests he would rather learn that process in house than hand control of a project this central to Tesla’s robotics and autonomy plans to an outside company.

Continue Reading

Elon Musk

Trump to hand Elon Musk a top honor that traces back to JFK

Trump will award Elon Musk the National Medal of Science at Thursday’s White House summit.

Published

on

By

elon musk and donald trump in front of a tesla cybertruck at the white house

Elon Musk is set to receive the highest honor the U.S. government gives to scientists and engineers.

President Donald Trump will present Musk with the National Medal of Science on Thursday at the White House’s Science: A New Golden Age Summit, Fox News Digital first reported on Wednesday. Google cofounder Sergey Brin, Nvidia CEO Jensen Huang and AMD CEO Lisa Su will receive the same medal, while Dell Technologies CEO Michael Dell and Microsoft CEO Satya Nadella will receive the National Medal of Technology and Innovation. A White House official later confirmed the list to Reuters.

“The Trump administration is grateful for the contributions of these incredible leaders in science and technology. These recipients are helping ensure America keeps leading the world in innovation,” White House spokesperson Liz Huston told Fox News.

It will be the first time Trump has presented either medal in his two terms. Congress created the National Medal of Science in 1959, and the National Science Foundation, which administers it, says 529 scientists and engineers have received it since. A presidential committee reviews nominees, but the president makes the final call.

Thursday’s group of medalists run or founded companies, and three of them sit at the center of the Super Intelligence hardware race that Musk competes in. Huang’s Nvidia supplies the GB300 chips filling SpaceX’s Colossus 2 cluster, while Su’s AMD is Nvidia’s biggest rival in data center GPUs.

Worth noting that Trump’s uncle, MIT physicist John G. Trump, received the National Medal of Science from President Ronald Reagan for his work on ionizing radiation and its uses in medicine and industry.

The Pentagon taps Elon Musk to design the battlefield of the future

For Musk, the medal is the latest sign of how far his relationship with Trump has come since their 2025 split over the “Big Beautiful Bill” and his exit from DOGE. Last week, he sat at Trump’s left during a White House lunch where AI executives signed a voluntary safety accord, and Defense Secretary Pete Hegseth named him to help lead the Pentagon’s Project Meridian study on the future of warfare. Musk has also adopted the administration’s new vocabulary, saying on Sunday that SpaceXAI will be renamed SpaceXSI after Trump ordered federal agencies to replace “artificial intelligence” with “super intelligence.”

Musk has collected science honors before, including the Stephen Hawking Medal for Science Communication in 2019.

Continue Reading

Lifestyle

Tesla FSD changed its mind mid-intersection, and it may have saved a life

Tesla shares dashcam footage of FSD Supervised stopping mid intersection to avoid a T-bone crash.

Published

on

By

Credit: @BLKMDL3/X

Tesla is putting another Full Self-Driving save in front of its 24.8 million followers on X.

On Tuesday morning, Tesla’s main account shared a dashcam clip with the caption “FSD Supervised preventing T-bone crash.” The footage came from an owner posting as TheNewGrid, who described what happened at a stop sign: “I looked at the car coming to the stop sign figured they would stop, my car went, then came to a stop mid intersection as they flew by. Had I been manually driving this would have resulted in a crash.”

The sequence is the notable part. FSD had already started crossing when the other driver ran the stop sign. Instead of pressing on, the car braked hard in the middle of the intersection and let the crossing vehicle pass in front of it. By the owner’s own account, they had made the same assumption the software initially made, that the other car would stop, and would not have corrected in time.

The clip is the latest in a run of safety posts Tesla has amplified over the past several days. On Saturday, the company shared a video from Selling Sunset star Jason Oppenheim, who sold his Bentley for a Model Y and said he was buying Teslas with FSD for 10 of his employees. Ashok Elluswamy, who leads Tesla AI, followed up by writing that Tesla self-driving “reacts to other people cutting into your path with super-human response times.” On Monday, a Cybertruck owner posted footage of FSD moving across three lanes from a red light to clear a path for an ambulance approaching from behind.

This recent clip also lands a few weeks after Tesla began shipping Automatic Collision Evasion with FSD v14.3.9, a feature that can activate FSD on the driver’s behalf when a frontal collision is imminent or the driver appears distracted. Elluswamy said in September that “even earlier prediction of hazards, even faster reaction time and overall significantly better safety and collision avoidance” are coming with v15, the release Tesla has tied to round the clock Robotaxi operation.

The safety messaging matters beyond social media. Tesla has said FSD Supervised was 4.1 times less likely to crash than manual driving across 100 million kilometers on European roads, and it has been putting those figures in front of regulators. Eight EU countries have now approved FSD Supervised, with Croatia the most recent, but the EU’s bloc-wide vote originally set for October 6 has been pushed to December at the earliest.

FSD Supervised is still a Level 2 system, and the driver remains responsible at all times. Even heavy users find reasons to step in. Teslarati’s Joey Klender, who uses FSD for about 76 percent of his driving, laid out five recurring issues on Tuesday that still prompt him to intervene. Clips like this one show the other column of that ledger: moments where the software caught a mistake a human was about to make.

Continue Reading