SYSTEMS

Startup wants to take atomic power generation on a barge to wherever the need is

Startup Bluecore Energy has raised $50 million in seed funding to develop floating nuclear power plants intended to supply ports and other locations with rapidly growing energy demands.

The California-based company said the Silverton Partners-led round, which includes $10 million in previously announced pre-seed funding, brought in several new investors. It plans to use the money to engineer and test its compact nuclear power system and pursue regulatory approval and maritime classification.

Its initial design is intended to generate about 10 MWe – megawatts of electrical output – continuously using light-water reactor technology. Bluecore says the maritime system would be able to operate for years at a time and require refueling only once every few years.

The company has secured its first barge and is developing a non-fueled, electrically heated reactor module prototype at its Port of Long Beach headquarters to test monitoring, sensor, and control systems.

Bluecore is targeting ports, AI datacenters, and coastal infrastructure, with the reactors potentially deployed offshore and connected to customers by subsea cables. Remote islands and coastal settlements in places such as Alaska may also lack practical connections to a wider electricity grid.

Bluecore claims its approach could provide reliable, zero-emission power to energy-intensive locations and open additional markets for nuclear generation.

“Six months ago, we were building the foundation. Today, we have the capital, the team, the hardware, and some of the most important institutions in nuclear and maritime working alongside us,” said Bluecore founder and CEO Kofi Asante.

“Our focus now is simple: create and deliver zero-emission energy as safely and quickly as possible.”

The move follows the launch of the International Atomic Energy Agency’s (IAEA) Atomic Technologies Licensed for Applications at Sea (ATLAS) initiative in the US capital last month.

Secretary of Energy Chris Wright linked that event to the Department of Energy’s own efforts to rapidly expand the provisioning of atomic energy generating capacity in order to meet the requirements of AI datacenters and other industries.

Bluecore has begun discussions with the US Nuclear Regulatory Commission (NRC) and Coast Guard. The two agencies recently signed a memorandum of understanding to coordinate oversight of the design, construction, and operation of civilian floating nuclear power plants. The NRC says it is ready to license maritime reactors under existing federal regulations, while it develops an additional framework for microreactors and similarly low-risk designs.

Bluecore remains years from commercial deployment. It has yet to build a fueled reactor or apply for the approvals required to operate one, and its Long Beach proposal must also navigate California’s restrictions on new nuclear power plants.

While some may have concerns about waterborne atomic power plants, America has a long history of maritime nuclear generation. All of the US Navy’s active aircraft carriers are nuclear-powered, but the technology first went to sea aboard the submarine USS Nautilus in 1955. USS Enterprise became the first nuclear-powered aircraft carrier in 1961, while the cruiser USS Long Beach, commissioned the same year, was the first nuclear-powered surface combatant.

Bluecore says its team includes former US Navy nuclear submarine officers and leaders with experience across organizations including SpaceX, Northrop Grumman, Toyota, Rivian, and Uber.

Russia currently operates the world’s only floating nuclear power plant. The barge-like Akademik Lomonosov is considerably larger than Bluecore’s proposed design and can generate up to 70 MWe. ®

• South Korea could invest $22.3 billion in a massive Texas gas plant
• The proposed facility would generate 6.3 GW of electricity for AI
• Seoul says negotiations remain unfinished, leaving the reported investment unconfirmed

South Korea and the United States have reportedly settled on how much Seoul will contribute towards a new energy facility in southern Texas.

Korean outlet Edaily, citing unnamed officials and lawmakers, reported the contribution at close to $22.3 billion for the venture, which would be built in Encinal, Texas, and is expected to produce 6.3 gigawatts of electricity to power AI data centers once fully operational.

Nintendo held its The Legend of Zelda 40th Anniversary Direct on Tuesday to reveal a new release for one of its biggest franchises. The big news was the reveal of The Legend of Zelda: Ocarina of Time remake release date or a special edition Nintendo Switch 2. But the gaming giant also announced an actual ocarina that fans can buy to annoy their friends and family while learning how to play.

The Legend of Zelda: Ocarina of Time remake will launch on Nov. 5, according to Tuesday’s presentation. A week before its release, a special edition Switch 2 for the 40th anniversary will be released on Oct. 29, along with a Switch 2 Pro Controller and carrying case. Then, to complete the anniversary package for Zelda fans, Nintendo plans to release an official ocarina identical to the one Link plays throughout Ocarina of Time. There will be a traditional acoustic model and an electric version made by Hori.

The main star of Tuesday’s direct was the Ocarina of Time remake. As expected, it will make full use of the Switch 2’s graphics power to give a modern look to the almost 30-year-old game, first released for the N64. Visuals weren’t the only upgrades, as controlling Link is much easier using both analog sticks, which weren’t around in 1998, and instead of jumping automatically near any ledge, there’s a jump button to let Link get across gaps.

Another interesting feature involves the ocarina Link uses in the game. This is a key item as it lets Link do various magical feats when a certain tune is played by pressing the correct buttons. Something different is that players can now hum the tune, and the game will recognize the sound via the Switch 2’s microphone and complete the input to activate the song.

Only the intro sequence of the game in Kokiri Forest and the first section of Hyrule Castle Town were shown during the presentation. All the characters received a visual upgrade as well as voices, with new fully voiced cutscenes. Link, however, is still not going to talk.

First-time players who get lost or a little confused in the game will be able to check out Zelda Notes, a new feature added into the Nintendo Switch app that lets players track their progress, gives some hints on in-game location and provides other info for players.

Before playing Ocarina of Time remake, fans can pick up the Nintendo Switch 2: The Legend of Zelda 40th Anniversary Edition system. The new Switch 2 variant will have a unique look on the console and dock inspired by the game. There will also be a Switch 2 Pro Controller and a carrying case with their own inspired looks that will be sold separately. Nintendo also plans to have an exclusive Switch 2 Pro Controller with its own display stand for sale only at the Nintendo Store in New York City and San Francisco.

Nintendo

Then, there’s the ocarina. Two ocarinas will be available for purchase: a standard acoustic ocarina and an electric one made by Hori that can play songs via a button. Eiji Aonuma, the producer of the Zelda series, gave a demo of how the ocarina sounds during the show. Nintendo didn’t provide a price or release date for these ocarinas, but it said in a press release that they will be available later this year.

Nintendo/Screenshot by CNET

Other Zelda news mentioned in the show was the April 30, 2027, release date for a film based on the game, officially called The Legend of Zelda. A 40th anniversary concert series featuring the game’s music will also kick off in January in Japan and in February in the US, starting in Chicago. Also out next year are the Young Link and Young Zelda amiibo figures.

Nintendo didn’t officially put a price on the Ocarina of Time remake, but the digital version is expected to be $60. There is a listing for the physical version of the game on the Nintendo Store for $70 that appeared shortly after the presentation and was taken down by Nintendo. It’s unclear what items shown during the event will be available at other retailers and which ones will be available only through the Nintendo Store.

Oscar Gonzalez

Contributing writer

Oscar Gonzalez is a Texas native who covered video games, conspiracy theories, misinformation and cryptocurrency.

See full bio

Advertisement

Xiaomi, at least in China, makes some of the best smartphone accessories, and power banks are definitely at the top. And after giving Indian consumers access to its latest air purifiers, Xiaomi has just expanded its portable charging lineup in India with three new power banks. The new range includes the Xiaomi UltraThin Magnetic Power Bank 5000, Xiaomi Magnetic Power Bank 10000, and the flagship Xiaomi Power Bank 20000 165W.

While the two magnetic models focus on wireless convenience and portability, the 20,000mAh variant is designed for users who need to fast-charge laptops, phones, and multiple devices simultaneously. Prices start at ₹3,699, with sales beginning on September 11.

Xiaomi UltraThin Magnetic Power Bank 5000 is Just 6mm Thick

If portability is your top priority, Xiaomi’s new UltraThin Magnetic Power Bank 5000 is the standout option. The company claims it’s only 6mm thick and weighs 98g, making it slim enough to slip into a pocket without adding much bulk.

It packs a 5,000mAh battery and supports 15W magnetic wireless charging for compatible devices. For faster charging, the USB-C port can deliver up to 22.5W. It can also charge two devices at once, combining wired and wireless charging.

On the other hand, the mid-range model is aimed at travelers and heavy smartphone users who need more capacity without sacrificing magnetic charging.

It features a 10,000mAh battery, 25W Qi 2.2 wireless charging, and a built-in USB-C cable so you don’t need to carry one separately. Wired charging reaches 45W, while support for PD, PPS, and QC 2.0 ensures compatibility with a wide range of phones and accessories.

Xiaomi Power Bank 20000 165W Can Fast-Charge Laptops

The flagship model is built for power users. With a 20,000mAh battery and 165W combined output, it can charge multiple devices at the same time, including compatible laptops.

USB-C 1 delivers up to 120W, while USB-C 2 supports 45W. There’s also an integrated 120W USB-C cable and a USB-A port for additional devices. A built-in color digital display shows battery percentage, charging speed, estimated time remaining, and system status.

Xiaomi says the power bank supports major fast-charging standards, including PD 3.0 (PPS), QC 3.5, BC 1.2, and Apple 2.4A.

from the say-it-loud dept

This week, our first place winner on the insightful side is Thad with a comment about calls for RFK Jr.’s resignation coming entirely from Democrats:

And that’s why trying to distill partisan politics down to just “team blue” vs. “team red” is reductive at best and false equivalence at worst. This and many other life-or-death issues break down on pretty clear partisan lines, and it’s entirely clear which party is right and which one is wrong.

Christ knows the Democrats aren’t perfect. But of the two major parties they’re the one that’s not a eugenicist death cult.

In second place, it’s an anonymous comment on our post about monitor makers hitting owners with annoying ads, and our characterization of how it happened:

No, that’s not how computers work. Microsoft Windows identifies the connected monitor, and then, on its own authority, checks whether the manufacturer has “recommended” any software to go with it; if so, Windows installs that without telling the user. The monitor isn’t installing anything.

Of course, we should also be criticizing monitor manufacturers for taking advantage of this misfeature. But the root problem is with Microsoft. More than 2 decades after the Sony rootkit, they haven’t learned their lesson (or maybe they have: people will put up with this shit, and the blame will mostly be misdirected).

Advertisement

For editor’s choice on the insightful side, we’ve got a pair of comments from. MrWilson. First, it’s one about giving Elon Musk credit when his products work:

Nope. Not at all. Screw that noise.

Teslas might function as a car some of the time, but that doesn’t mean you have to give Musk any credit whatsoever.

Starlink is a money pit from whence nothing can be recovered. Even the most corrupt broadband fiber deployment can be dug up or reused. Comcast could, in a just judicial system, be broken up as an illegal monopoly and their infrastructure that we paid for could be reused by other service providers.

And not every dollar was going to go to big monopolistic companies, whereas any money going to Starlink funds literal authoritarian, anti-democratic election interference.

Advertisement

Next, it’s one on our post about the “Darnella Test” for social media regulation:

Unfortunately, “kids won’t be able to post videos that make authorities look bad and fuel significant protests” is an appealing argument to the authoritarians who are already in favor of curbing social media access for kids which those authoritarians can’t directly control.

Over on the funny side, things are extremely quiet this week, so we’re going to truncate it and just feature the only comment that received enough votes to earn a funny badge. It’s Stephen T. Stone with a reply to a comment calling Mitch McConnell a vegetable:

How dare you insult vegetables like that. Unlike Mitch McConnell’s political career, vegetables are actually good for the average person.

That’s all for this week, folks!

Apple will host its annual iPhone event on Wednesday, September 9 at 10 a.m. PDT, and it’s expected to be a big one. All signs point to Apple’s long-awaited entrance into the foldable smartphone market, finally shipping this kind of hardware about seven years after Samsung did.

Apple has a YouTube stream prepared for the “Surprise and shine” event, and you can set reminder notifications within YouTube to make sure you don’t miss it (or you can bookmark this page). You can also watch the livestream on Apple’s own events site.

Aside from the foldable iPhone Ultra, Apple is also expected to announce the iPhone 18 Pro and iPhone 18 Pro Max models, for people who want an upgrade but don’t see the appeal of the foldable, or who don’t want to pay the rumored $2,000 price for the iPhone Ultra. The standard iPhone 18 model isn’t expected to launch until next year.

Apple may also release new hardware like an updated HomePod, camera-enabled AirPods, and new Apple Watches.

With new Apple CEO John Ternus at the helm, this will be Apple’s first major product announcement since longtime CEO Tim Cook stepped down.

I’m sitting in a Rivian R1S SUV as it drives itself down the leafy streets of Palo Alto, Calif., through areas crowded with touchstones of tech history. At one point I pass the landmark HP Garage, the one-car workshop where Hewlett-Packard, and, arguably, Silicon Valley, was founded in 1939. I skirt Stanford University, where a team led by computer science professor Sebastian Thrun won a US $2 million DARPA Grand Challenge in 2005. The team’s Volkswagen SUV, named Stanley, became the world’s first vehicle to navigate a grueling 212-kilometer Mojave Desert course with no human intervention.

The Rivian I’m in might look like any other electric SUV in this affluent town, with its concentration of tech bros, venture capital, and startups. But inside this boxy EV is something special: an Autonomy+ system that will allow owners to enter an address, sit back, and let the vehicle drive to any mapped destination in the U.S. and Canada. This point-to-point system is one of the most advanced semiautonomous-driving systems coming to market. It is also a precursor of the company’s bid to make self-driving cars a reality, for robotaxis and—eventually—for everyday car buyers. After years of incremental advances and frustrating setbacks, self-driving has been swept up in the great AI resurgence, and is now a top priority for investors and global automakers, who envision vast new streams of profits.

So here I am, 21 years after that DARPA challenge, riding shotgun in Stanley’s vastly more advanced descendant. Rivian’s Autonomy+ is intended to operate seamlessly on suburban streets like these, sensing and responding to traffic lights, crosswalks, and stop signs. That point-to-point system is set to debut on Rivian’s all-new R2 SUV by roughly the end of this year, and via over-the-air updates for its newest R1S and R1T models. Rivian says it will charge $49.99 a month, or $2,500 up front, versus Tesla’s $99 per month for its rival system, which is somewhat misleadingly called Full Self-Driving (Supervised), or FSD. Mercedes, meanwhile, plans to charge $3,950 for a three-year subscription for the forthcoming MB.Drive Assist Pro on its CLA-Class EV; that system still requires at least one hand on the steering wheel.

Video released by Rivian shows the company’s R1 SUV being driven on a variety of urban and rural roads, according to the company. Rivian plans to introduce this self-driving system to compete with Tesla’s offering before the end of 2026. Rivian

Impressive as it is, Autonomy+ is only a Level 2+ system in the classification system established by the Society of Automotive Engineers. Level 2+ means that a human driver must be ready to retake control at any moment. Rivian, along with a horde of deep-pocketed rivals, is aggressively working toward more impressive (and potentially lucrative) levels of autonomy. At Level 3, drivers could “check out” behind the wheel for limited periods, to scroll through emails or watch a movie—but not to sleep.

The big race right now is to deliver Level 4 autonomy: A car you could (in theory) dispatch to pick up a pizza, and have it carted home on the heated, unoccupied driver’s seat—or in which you could spend the ride lounging alone in the back seat, enjoying a private slice while reading a newspaper.

At Rivian’s software lab in Palo Alto, Calif., a technician evaluated code for the company’s self-driving system.Jason Henry/Bloomberg/Getty Images

Robotaxis currently roaming the U.S., China, and the Middle East have proved that driverless, Level 4 autonomy is possible. These cars operate in relatively tiny numbers in a couple of dozen cities, and within the specific constraints of commercial services. Now Rivian and its many rivals—including Tesla, Toyota, Mercedes, Volkswagen, and China’s BYD are racing to bring that level of self-guided mobility to the masses. Rivian’s strategy combines a suite of cameras, radar, and lidar; a custom set of silicon chips, developed in-house, to process sensor data; and an AI autonomy model running on those chips. With $1.25 billion in backing from Uber, Rivian plans to graduate to a fleet of self-driving, Level 4 robotaxis starting in 2028. Those taxis, in turn, will be the literal training wheels for extending Level 4 ability to consumer vehicles.

Meanwhile, millions of connected cars, as they cruise every nook and cranny of the globe, are already sending data to train automakers’ systems. The race is on to funnel those data through fast-improving AI models with “end to end” capability: an AI architecture, powered by deep learning, that processes raw sensor data directly into physical vehicle commands. So equipped, engineers anticipate they’ll be able to solve the tricky edge cases—tangled urban streets, unique geographies, swarms of pedestrians, inclement weather—that skeptics once deemed intractable.

Rivian’s Plan for Level 4 Self-Driving

Despite the company’s high media profile, including a spotlight on RJ Scaringe, its MIT-doctorate founder and chief executive, Rivian holds a relatively tiny slice of the U.S. passenger-vehicle market. It sold just 42,000 vehicles last year across its three models, the adventure-minded R1S SUV and R1T pickup, and the Electric Delivery Van. Tesla sold about 1.6 million units. Toyota, the world’s largest automaker, sold more than 11 million.

The first generation of the Rivian Autonomy Processor, an AI processing chip developed in-house, was tested at Rivian’s Palo Alto, Calif., lab in December, 2025. Jason Henry/Bloomberg/Getty Images

Rivian’s underdog strategy is to leverage software and tech to make itself a serious player. Volkswagen, among the world’s largest automakers, saw enough value there to invest up to $5.8 billion in a joint venture called Rivian and Volkswagen Group Technologies. The joint venture gives Rivian crucial capital for development. It gives Volkswagen access to Rivian’s electrical architecture and to the software for the R2, new-generation Rivian SUV that went on sale in June.

Unlike traditional lidar units, which protrude like a layer cake from the roof of a vehicle, Rivian’s unit on the new R2 SUV is housed in a small, sleek enclosure where the windshield meets the roof.Rivian (2);Jason Henry/Bloomberg/Getty Images

“Rivian developed an architecture so important that VW is spending billions to buy it, as opposed to trying to re-create it themselves,” says Bryan Reimer, a research scientist in MIT’s Center for Transportation and Logistics.

But the joint venture doesn’t give VW access to Rivian’s autonomous tech. In March, that R2 architecture underpinned Rivian’s $1.25 billion deal to supply Uber with up to 50,000 robotaxis. The companies plan to initially deploy 10,000 taxis, beginning in San Francisco and Miami in 2028, before expanding across 25 cities in the U.S., Canada, and Europe.

Rivian’s vulnerabilities include struggles with reliability, along with expensive body repair costs that the company says it strove to reduce for its new R2. As impressive as Rivian’s in-house tech may appear, the company has miles to go to catch up with Tesla, which recently announced it has 1.1 million active users of its FSD system. Toyota is also jumping into the game; its Woven by Toyota subsidiary has partnered with the Alphabet-owned Waymo to develop an autonomy platform for robotaxis and consumer cars.

The lidar unit on a Waymo robotaxi protrudes noticeably from the roof of the vehicle.Andrej Sokolow/picture alliance/Getty Images

Until recently, most observers would have gone all-in on Tesla as the winner of the autonomous race. Elon Musk’s company has begun operating a small test fleet of Model Y robotaxis in three Texas cities and in Florida. Tesla has also begun producing a dedicated autonomous vehicle, the Cybercab robotaxi. But in April, Musk pushed back his timeline for Level 4 autonomy for general consumers: “I’m just guessing here, but probably in the fourth quarter” of 2026, he said. It was the latest in a series of deflating walkbacks from the man who once promised 1 million robotaxis on the road by 2020.

Scaringe, during an unveiling of his company’s make-or-break R2 SUV at a Utah state park, says that showroom Rivians will start adopting some of its robotaxis’ Level 4 capabilities no later than 2030, perhaps beginning with self-parking functions.

How Self-Driving Systems Are Learning From Humans

Like most autonomous cars, Rivian’s system fuses data from multiple sensors to create a robust picture of a fast-moving environment and its obstacles. Data is fed to a neural network—what Rivian refers to as its “Large Driving Model,” or LDM—that churns through hundreds of trillions of operations per second to interpret and fuse data from cameras, radar, and lidar. That network is end to end, meaning that it processes multiple streams of raw sensor data (such as camera pixels) and outputs driving controls (for steering, braking, and acceleration) through a single data pipeline. More traditional systems coded distinct steps for data collection, feature extraction, prediction, and decision-making.

That proprietary AI driver identifies features in images and point clouds, groups them into objects, and tracks them across frames, time-stamped to the millisecond to account for differing frame rates. The AI thus builds confidence over time, acting on object detections that persist across several frames, rather than, say, slamming the brakes due to a camera blip on a single frame. The virtual driver can then navigate safely even when sensors disagree, by favoring the persistent data. The output— commands for electric motors and other systems—is backed by redundant hardware for by-wire systems such as steering and brakes.

During my demo of Rivian’s point-to-point Autonomy+ system, a company test driver sits behind the wheel. Nick Nguyen, the engineer who directs Rivian’s products and programs related to autonomy, watches from the back seat. Compared to, say, a large language model that writes news or fiction, Nguyen says, an autonomous-driving AI is easier to evaluate, so there’s little room for error. “We want cliché. We want boring. Just safe, repeatable driving,” he says.

The Rivian R2 is a mid-size SUV with self-driving and off-road capabilities. It competes with the more urban-oriented Tesla Y. Rivian

From my brief drive, I’d say suburban boredom is achieved in this Rivian R1S. Unlike some modes of Tesla’s Full Self-Driving (Supervised), Rivian’s system drives like a soccer dad, obeying speed limits to the digit, stopping gracefully at traffic lights, and easing over speed bumps like a driver delivering antiques. For robotaxi companies in the U.S. and China, these types of ho-hum trips are boosting optimism and investment to dizzying heights. Waymo claims 92 percent fewer fatal or serious-injury accidents than human drivers, based on 170 million miles of autonomous ride data. But the real challenge is how well the higher levels of autonomy will work when they reach consumer cars [see Sidebar, “The Growing Proof That Autonomous Cars Save Lives”].

Rivian’s core LDM currently ingests cloud data from up to 125,000 cars for analysis and validation, which then fine-tunes the model through simulations. Onboard computing is smart enough to trigger recording only for unusual scenarios. Owners have to agree explicitly to data collection beforehand. Updated LDMs will be beamed back to customer cars via monthly over-the-air updates, part of that self-reinforcing data flywheel.

As is true for some of its rivals, Rivian no longer needs to equip its vehicles with an onboard high-definition map or even a cellular link as a backup to pinpoint the car for navigational purposes. That strategic shift reduces data demands, and ensures steady driving in urban canyons or tunnels with no connections. Instead, the Rivian recognizes and responds to its surroundings through recognition and repetition, just as a human would do (and also just as Tesla’s FSD does): interpreting street signs, following lane markers, being alert to hazards. The Rivian R2 features 11 high-definition cameras and five radars. It will integrate a lidar unit early next year to lay the groundwork for future autonomy. That miniaturized lidar will integrate smoothly into the R2’s existing roofline, an improvement over the bulky, drag-producing units seen on Waymo Jaguars, and older partially autonomous models. Vidya Rajagopalan, Rivian’s senior vice-president of electrical engineering hardware, says lidar costs have fallen from above $10,000 to a few hundred dollars in under a decade.

Vidya Rajagopalan, Rivian’s senior vice president of electrical engineering hardware, holds a RAP1 AI processor chip.Jason Henry/Bloomberg/Getty Images

A mix of sensors plays up the strengths and diminishes the weaknesses of each, Rajagopalan says. Cameras capture color and texture and can distinguish between objects, but they struggle in darkness and low-contrast lighting. Lidar is unaffected by darkness or blinding sunlight, and senses shapes in three dimensions. This inherent 3D capability makes lidar more reliable for slowing or halting a car for random objects—“a tire in the road, or maybe a large dinosaur,” Nguyen quips. Multiple cameras can further contribute 3D data, after a short delay for processing.

Sensors with 360-degree vision can outperform human senses in key situations. Radar and lidar can spot nighttime pedestrians or animals hundreds of meters down the road, something no human can do. But lidar can be thrown off by dust, fog, and snow. Radar can “see” through rain or snow, but with relatively low spatial resolution.

Why Rivian Ditched Nvidia

To handle the flood of sensor data, Rivian has taken on an ambitious challenge: designing its own custom autonomy chip in-house. The Rivian Autonomy Processor (RAP1) is a 5-nanometer processor that can execute 800 trillion operations per second (TOPS), three times as fast as the Nvidia Jetson Orin chip used in its earlier models. The chip will be built to Rivian’s specs by Taiwan Semiconductor Manufacturing Co. , which also makes custom chips for Tesla.

Rivian’s autonomy module contains two Rivian Autonomy Processors, each capable of 800 trillion operations per second.Rivian

Nvidia’s latest automotive system-on-a-chip, the Drive AGX Thor processor, is being adopted by the likes of BTD, Hyundai, Lucid, Mercedes, Nissan, Volvo, and Xiaomi, along with the Aurora and Waabi autonomous-trucking companies.

On paper, a single AGX Thor chip is slightly faster in terms of TOPS, at 1,000 trillion operations per second. But Rivian combines a pair of chips in each autonomy module, giving it 1,600 TOPS and execution rates around 5 billion pixels of data per second, versus 3.5 billion for Nvidia’s Thor.

Rajagopalan says developing the chip and AI software simultaneously shaved a critical full year from development. Experts say it’s the kind of fast-to-market speed that China has mastered and that legacy automakers are struggling to match. The in-house design allows Rivian to custom-tailor its software to the chip, and vice versa. Nvidia’s general-purpose chip, designed to satisfy multiple customers with various needs, must devote computing power to onboard infotainment, displays, or other systems. Rivian’s chip is designed to run autonomy and nothing but.

During my visit to Rivian’s Silicon Valley campus, Rivian engineers Prasun Raha and Mukund Chavan tutored me on the rapid pace of the company’s autonomy evolution. A cluttered wallboard displays a first-gen architecture that Rivian debuted just five years ago. The initial R1S SUV and R1T pickup used nearly a score of electronic control units (ECUs), the “black boxes” that traditionally control vehicle functions. For its latest R1 models, Rivian reduced the ECU count to seven. The zonal architecture organizes nearly every vehicle function into three zones, hugely consolidating the electronics and simplifying manufacturing. Rivian also leaned into an autonomy trend called “early fusion”: mixing raw, time-and-space-aligned sensor data into a shared view before the neural network acts upon it. In late fusion, each sensor performs solo recognition before it’s combined into a single picture.

The self-driving revolution will really begin when the technology migrates from controlled taxi fleets to consumer cars, giving owners back the precious time they waste on commuting.

Early fusion preserves the richest sensor data for maximum accuracy in self-driving. But it demands the enormous computing power the RAP1 can deliver. Raha says the approach helps the multimodal system degrade gracefully and continue to operate with certainty even if, say, a camera’s lens gets covered with mud.

Together, these elements make up Rivian’s third-generation autonomy platform. Displayed on a test bench, a new Autonomy Compute Module pairs two RAP1 supercomputing chips. The module is eight times as powerful as before but 60 percent smaller, according to the company. Raha says the system was designed expressly to expand Rivians to Level 4 autonomy from today’s Level 2+. RivLink, the automaker’s interconnect technology, can bridge multiple RAP modules to scale processing power. “It lets us build this extensible system, with perhaps two more chips for Level 3 or four for Level 4, depending on how the model scales,” Raha says.

Rivian’s Road Map to Full Autonomy

Rivian’s next planned milestone toward self-driving will be Level 3 autonomy—a hands-off and eyes-off system, but for highways only. (Remember, Tesla’s current FSD is technically a Level 2 system: hands off but not eyes off.) On the freeway, Nguyen points out, drivers would be spared the drudgery of dealing with stop-and-go traffic, allowing them to boost productivity or just goof off.

Some autonomy critics are leery of Level 3, envisioning a limbo zone in which drivers are lulled into a false sense of security when a car drives for long stretches with no human attention required. Ford and GM are among the automakers pivoting toward limited eyes-off functions.

Rivian’s senior vice-president of autonomy, James Philbin, sees Level 3 as an inevitable stepping-stone to Level 4. The company expects it will initially be limited to highways, not the cut-and-thrust of city traffic. If a driver fails to respond to alerts, the system will slow the vehicle, pull off on a shoulder, or call 911. Rivian has not announced a timeline for this Level 3 system.

Navigating a Tricky Liability Shift on the Way to Immense Profits

Ready or not, these much more autonomous systems are coming, a natural evolution of today’s semiautonomous helpers. In developed markets, adoption of showroom cars with partial-to-full automation is projected to jump from 8 percent in 2024 to 28 percent by 2030, according to Morgan Stanley.

“One in four cars sold globally may be equipped with smart-driving technology in five years, versus one in eight cars now,” wrote Tim Hsiao, a Morgan Stanley analyst, in a note posted on the company’s website.

Combining cameras, lidar, and radar gives a self-driving car a better view of people and objects in front of it, according to Rivian. The company expects to release a self-driving system before the end of 2026 that will compete with Tesla’s, which uses cameras alone. Rivian

MIT’s Reimer believes the self-driving revolution will really begin when the technology migrates from controlled taxi fleets to consumer cars, giving owners back the precious time they waste on commuting. If owners could truly send their autonomous car to safely chauffeur children, keep aged parents mobile, or run errands—while owners keep working or playing—the automakers who first help make that happen will enjoy a massive competitive edge, he says. As automakers struggle to convert buyers to subscription models, self-driving appears to be the one advance for which consumers might actually pay plenty.

But the greatest impediment to that revolution has little to do with technology. Public skepticism over self-driving is rampant; and the fate of fully autonomous testing in New York City is uncertain. Even going from Level 2 to Level 3 might shift legal liability for accidents in some cases to automakers from drivers, but with Tesla still fighting lawsuits over its rudimentary Autopilot systems, those questions aren’t anywhere near settled.

Experts worry that self-driving cars may become as politicized as EVs. Labor unions are pushing back, fearing job losses from taxis to trucking. A crazy quilt of state or local regulations has failed to create coherent industry guidelines. Publicized failures—even ones that don’t result in injuries, such as Waymos driving onto a flooded street or impeding emergency workers—give the industry a black eye. Companies like Tesla and even Waymo, Reimer says, have too often relied on an arrogant “Trust me” approach, resisting regulation and oversight.

Nevertheless, the momentum toward real, Level 4, eyes-off, autonomy has reached a point from which there’ll be no backing off. The rest of the journey will depend as much on social and regulatory issues as technical ones, and so Reimer has a bit of advice.

“Do it right, and share all your data,” he says. “Earn the right to scale…. It’s about establishing trust, and developing a framework in which we truly believe these systems can operate as a trusted part of our transportation network.”