Elon Musk's Terafab Massive Risks, Tesla Robotaxi Cybercab Production Hell? Here's Why!

Kind: captions Language: en So broadly speaking, we expect to make uh two two kinds of chips. One will be optimized for uh edge inference. So that'll be used primarily in Optimus and in the cars, but but especially in Optimus because I expect uh the robots uh humanoid robots to be made uh 10 to 100 times more than the volume of cars. The Terrafab launch event just happened and it's been a long time since Elon Musk got back on stage to talk about something this far beyond what most of us can even imagine. Just picture the scale for a second. The entire US power grid is only about 0.5 terowatt. But Terrafab is aiming to produce over 1 terowatt of computing every year. Basically double the country's total electricity capacity. And this factory is dedicated solely to making chips for AI, robots, and space exploration. Elon Musk called it the greatest chip building experiment in history. An effort to go beyond anything any chip factory in the world has ever done. That said, a lot of people think Terrafab could end up being one of the riskiest and most failureprone projects ever announced. Some even believe Roboaxi and Cyber Cab could fall into production hell because of it. So why is that? What's really going on behind this factory? What's inside it? And how could it impact Cybercap production? bigger than a terowatt. So, for sure the future I want to see I I want us to live long enough to see the the mass driver on the moon cuz that's going to be incredibly epic. >> Yeah. For those who don't know, Terapab is a joint venture between Tesla, SpaceX, and XAI, the AI company that SpaceX recently acquired through an all- stock deal. Terraab is not your typical semiconductor factory. It's a fully vertically integrated facility combining chip design, logic chip manufacturing, memory storage chips, and advanced packaging all under one roof. This facility is expected to be built in the north zone of Giga, Texas. Tesla's CFO admitted that the total cost of Terafab is estimated around 20 to $25 billion and has not yet been included in Tesla's record- setting capital expenditure plan for 2026, which already exceeds $20 billion. So, what exactly is Terraab's goal? The production goals are truly astounding. Terraab is designed to reach an initial output of 100,000 wafers per month with the ambition to scale up to 1 million wafers per month at full capacity. To put that into perspective, hitting full capacity would represent about 70% of TSMC's total global output, all from a single facility operated by companies that have never manufactured chips before. Musk stated that the facility will produce between 100 and 200 billion AI and custom memory chips per year, powering Tesla's full self-driving software, the CyberCap robot taxi program, and the Optimus humanoid robot line. He also mentioned that millions of Optimus robots will help build and operate the facility itself. As for why Tesla needs to make its own chips at the Terraab announcement event, Elon Musk acknowledged the contributions of current suppliers. But he also admitted that their production ramp up is too slow to keep up with Tesla's lightning fast growth. So to avoid being limited by third party production capacity, Tesla chose to build its own factory to ensure they have the necessary chips in the right quantity on time. Tesla said it's targeting 2nmter processing technology, the most advanced currently entering commercial production. TSMC has only just begun ramping up its own 2nmters output after decades and hundreds of billions of dollars spent building that capability. To simplify the complex process, Musk revealed that Terapab will focus on just two main types of chips and inference chip for Tesla vehicles and Optimus robots. Similar to Tesla's current AI4 and a custom D3 chip designed specifically for AI satellites in orbit, small-scale production of AI5 is expected to begin in 2026 with mass production planned for 2027. It's worth noting, however, that Tesla had already delayed the AI5 launch to mid 2027 prior to this announcement and the AI6 chip was pushed back about 6 months due to delays in Samsung's 2n chip manufacturing process. The [snorts] most surprising part of the presentation was the vision for space computing. Musk stated that 80% of Terrafab's compute output will be dedicated to AI satellites in orbit with only 20% allocated to groundbased applications. He explained that solar radiation in space is about five times stronger than on Earth and the vacuum environment makes thermal scaling much more feasible. His conclusion, AAI computing in orbit could become cheaper than terrestrial alternatives within the next 2 to 3 years. Musk describes Terapab as the final piece needed to unlock an era of abundance self-evolving AI. billions of humanoid robots and a multilanetary humanity. On closer analysis, Terrafab doesn't just solve the chip bottleneck. It also creates a competitive advantage that would be nearly impossible to replicate. From an economic standpoint, 20 to 25 billion is an enormous sum. But if successful, it could deliver cheaper, more independent compute and support Musk's entire ecosystem. However, many people remain skeptical of Elon Musk's claims, arguing that he may be repeating a familiar pattern from Tesla's battery day event in 2020, where a similar scenario unfolded. In September 20th, 20 Musk took the stage and promised a revolution in battery manufacturing with the 4680 cell. Tesla aimed to ramp production to 10 gawatt hours within a year and eventually reach 3 TWWs by 2030, enough for 20 million vehicles annually. The dry electrode process was expected to cut costs by up to 50%. However, after 5 and a half years, the 4680 program has been disappointing. Tesla's own leading battery supplier has stated that Elon doesn't know how to manufacture batteries. The dry electrode process required six or seven major revisions. The timeline stretched years beyond initial promises, and the 3TW goal now appears to be a distant dream. Estimates suggest Tesla has achieved only about 2% of its original battery production target. Now Musk wants us to believe he will build a chip fabrication plant. Not just any fab, but the largest in the world using 2 nanometers technology and producing 70% of TSMC's total output from a single building. Battery manufacturing is already difficult. Advanced semiconductor manufacturing is exponentially more challenging. As we know, TSMC has spent $165 billion over several years to build six fabs in Arizona, and those facilities are not expected to reach 2 nanometers production until 2029. A single 2nm fab with a capacity of 50,000 wafers per month costs around 28 billion, and construction alone in the US takes about 38 months. Tesla has no experience in semiconductor manufacturing. Morgan Stanley estimates that building large-scale chip production capacity would require total capital investment of roughly 35 to4 billion. The most groundbreaking highlight is the plan to move 80% of computing capacity into orbit, a space-based data center system powered by solar energy and deployed using Starship. While this vision is ahead of its time and challenges conventional business norms, it reflects an unlimited ambition to redefine the infrastructure of future technology. That's why we see this as a bold highstakes bet by Elon Musk by announcing an eighth generation AI chip even as earlier generations are still being finalized. He is betting on a leap in speed that even giants like TSMC would have to take seriously. What do you think about Terraab? Do you believe Elon Musk's claims? Drop a comment below. Thanks for watching and goodbye. [music]