Nvidia’s Cosmos Offers Synthetic Training Data; Following Tesla’s Lead

Not a Tesla App

At the 2025 Consumer Electronics Show, Nvidia showed off its new consumer graphics cards, home-scale compute machines, and commercial AI offerings. One of these offerings included the new Nvidia Cosmos training system.

Nvidia is a close partner of Tesla - in fact, they produce and supply the GPUs that Tesla uses to train FSD - the H100s and soon-to-be H200s, located at the new Cortex Supercomputing Cluster at Giga Texas. Nvidia will also challenge Tesla’s lead in developing and deploying synthetic training data for an autonomous driving system - something Tesla is already doing.

However, this is far more important for other manufacturers. We’re going to take a look at what Nvidia is offering and how it compares to what Tesla is already doing. We’ve done a few deep dives into how Tesla’s FSD works, how Tesla streamlines FSD, and, more recently, how they optimize FSD. If you want to get familiar with a bit of the lingo and the background knowledge, we recommend reading those articles before continuing, but we’ll do our best to explain how all this synthetic data works.

Nvidia Cosmos

Nvidia’s Cosmos is a generative AI model created to accelerate the development of physical AI systems, including robots and autonomous vehicles. Remember - Tesla’s FSD is also the same software that powers their humanoid robot, Optimus. Nvidia is aiming to tackle physical, real-world deployments of AI anywhere from your home, your street, or your workplace, just like Tesla.

Cosmos is a physics-aware engine that learns from real-world video and builds simulated video inputs. It tokenizes data to help AI systems learn quicker, all based on the video that is input into the system. Sound familiar? That’s exactly how FSD learns as well.

Cosmos also has the capability to do sensor-fused simulations. That means it can take multiple input sources - video, LiDAR, audio, or whatever else the user intends, and fuse them together into a single-world simulation for your AI model to learn from. This helps train, test, and validate autonomous vehicle behavior in a safe, synthetic format while also providing a massive breadth of data.

Data Scaling

Of course, Cosmos itself still requires video input - the more video you feed it, the more simulations it can generate and run. Data scaling is a necessity for AI applications, as you’ll need to feed it an infinite amount of data to build an infinite amount of scenarios for it to train itself on.

Synthetic data also has a problem - is it real? Can it predict real-world situations? In early 2024, Elon Musk commented on this problem, noting that data scales infinitely both in the real world and in simulated data. A better way to gather testing data is through real-world data. After all, no AI can predict the real world just yet - in fact, that’s an excellent quantum computing problem that the brightest minds are working on.

Yun-Ta Tsai, an engineer at Tesla’s AI team, also mentioned that writing code or generating scenarios doesn’t cover what even the wildest AI hallucinations might come up with. There are lots of optical phenomena and real-world situations that don’t necessarily make sense in the rigid training sets that AI would develop, so real-world data is absolutely essential to build a system that can actually train a useful real-world AI.

Tesla has billions of miles of real-world video that can be used for training, according to Tesla’s Social Media Team Lead Viv. This much data is essential because even today, FSD encounters “edge cases” that can confuse it, slow it down, or render it incapable of continuing, throwing up the dreaded red hands telling the user to take over.

Cosmos was trained on approximately 20 million hours of footage, including human activities like walking and manipulating objects. On the other hand, Tesla’s fleet gathers approximately 2,380 recorded minutes of real-world video per minute. Every 140 hours - just shy of 6 days - Tesla’s fleet gathers 20 million hours of footage. That was a little bit of back-of-the-napkin math, calculated at 60 mph as the average speed.

Generative Worlds

Both Tesla’s FSD and Nvidia’s Cosmos can generate highly realistic, physics-based worlds. These worlds are life-like environments and simulate the movement of people and traffic and the real-life position of obstacles and objects, including curbs, fences, buildings, and other objects.

Tesla uses a combination of real-world data and synthetic data, but the combination of data is heavily weighted to real-world data. Meanwhile, companies who use Cosmos will be weighting their data heavily towards synthetically created situations, drastically limiting what kind of cases they may see in their training datasets.

As such, while generative worlds may be useful to validate an AI quickly, we would argue that these worlds aren’t as useful as real-world data to do the training of an AI.

Overall, Cosmos is an exciting step - others are clearly following in Tesla’s footsteps, but they’re extremely far behind in real-world data. Tesla has built a massive first-mover advantage in AI and autonomy, and others are now playing catch-up.

We’re excited to see how Tesla’s future deployment of its Dojo Supercomputer for Data Labelling adds to its pre-existing lead, and how Cortex will be able to expand, as well as what competitors are going to be bringing to the table. After all, competition breeds innovation - and that’s how Tesla innovated in the EV space to begin with.

Follow

Follow

Follow

Tesla FSD in Europe: June Update

Not a Tesla App

The road to bringing FSD to Europe has been a long and complex one and filled with regulatory and bureaucratic hurdles. Elon Musk, as well as other members of Tesla’s AI team, have previously voiced their grievances with the regulatory approval process on X.

However, it appears that there is finally some progress in getting things moving with recent changes to upcoming autonomy regulations, but the process still seems slow.

Waiting on the Dutch

Elon commented on X recently, stating that Tesla is waiting for approval from Dutch authorities and then the EU to start rolling out FSD in Europe. Tesla is focusing on acquiring approvals from the Dutch transportation authority, which will provide them with the platform they need to gain broader acceptance in Europe. Outside of the Netherlands, Tesla is also conducting testing in Norway, which provides a couple of avenues for them to obtain national-level approval.

The frustration has been ongoing, with multiple committee meetings bringing up autonomy regulation but always pulling back at the last second before approving anything. The last meeting on Regulation 157, which governs Automated Lane Keeping Systems, concluded with authorities from the UK and Spain requesting additional time to analyze the data before reaching a conclusion.

Tesla, as well as Elon, have motioned several times for owners to reach out to their elected representatives to move the process forward, as it seems that Tesla’s own efforts are being stymied. 

This can seem odd, especially since Tesla has previously demoed FSD working exceptionally smoothly on European roads - and just did it again in Rome when they shared the video below on X.

FSD Supervised testing in Rome, Italy

Pending regulatory approval pic.twitter.com/TXh3uQIfrx

— Tesla Europe & Middle East (@teslaeurope) June 12, 2025

DCAS Phase 3

While the approval process has been slow, Kees Roelandschap pointed out that there may be a different regulatory step that could allow FSD to gain a foothold in Europe.

According to Kees, the European Commission is now taking a new approach to approving ADAS systems under the new DCAS Phase 3 regulations. The Commission is now seeking data from systems currently operational in the United States that can perform System-Initiated Maneuvers and don’t require hands-on intervention for every request.

This is key because those are two of the core functionalities that make FSD so usable, and it also means that there may not be a need to wait years for proper regulations to be written from scratch. Now, the Commission will be looking at real-world data based on existing, deployed technology, which could speed up the process immensely.

What This Means

This new, data-driven regulatory approach could be the path for Tesla to reach its previous target of September for European FSD. While the cogs of bureaucracy are ever slow, sometimes all it takes is a little data to have them turn a bit faster in this case.

Alongside specific countries granting approval for limited field testing with employees, there is some light at the end of the tunnel for FSD in Europe, and hopes are that a release will occur by the end of 2025. With Europe now looking to North America for how FSD is performing, Tesla’s Robotaxi results could also play a role.

Follow

Follow

Follow

Tesla Launches 'TeslaVision' Contest With Big Prizes — See Last Year’s Winner [VIDEO]

Not a Tesla App

Tesla’s marketing has always been relatively unconventional, relying on word-of-mouth rather than traditional advertising. The passion of the owner’s community is always massive, but it is especially high now with the launch of the Robotaxi network just around the corner.

Tesla is now tapping into that spring of fan creativity and announced the TeslaVision video contest, with some seriously impressive prizes up for grabs.

Announcing the TeslaVision fan video showcase

Tesla owners & supporters have always been able to see our products & mission for what they truly are.

Your word of mouth has made Tesla what it is today.

The OGs will remember that in 2017, we held our first video contest. 8 years… pic.twitter.com/6pPpkqmqOH

— Tesla (@Tesla) June 10, 2025

The Contest

The core of the contest is simple. Create a video that shows how your Tesla gives you more in life. Tesla is looking for submissions that highlight themes of freedom, safety, fun, and convenience.

Prizes

The prizes definitely make this contest worth entering if you’re good with a camera and have some basic video editing abilities.

For North America, the prizes include a brand new Model Y AWD Long Range, alongside an all-expenses-paid trip to Austin for a tour of Giga Texas. The grand prize winner will also be able to custom order their Model Y, allowing them to select their preferred wheels and color.

The two runners-up won’t get a Model Y, but they’ll also enjoy an all-expenses-paid trip to Giga Texas for a tour of the factory.

The travel and tour include lodging in Austin for 2 nights, as well as economy-class round-trip tickets from anywhere in North America. Tesla will also provide a vehicle for use during the trip.

Hopefully, these winners will also have the opportunity to experience the Robotaxi network while they’re in Austin, as it’s expected to be opened to the public later this month.

Project Loveday

For long-time followers of Tesla, this contest may feel familiar. The contest is a direct throwback to the 2017 Project Loveday contest, which was inspired by a letter to Tesla from a 10-year-old aspiring marketer. That contest was won by MKBHD, with his submission below:

How to Enter

If you’re ready to start filming, here are the key pieces of information you’ll need to know:

• Video must be 90 seconds or less

• Video must be uploaded to YouTube with a public URL

• Make a post on X and Instagram tagging “@Tesla” and include the words “TeslaVision contest” in the post.

• Provide links to both social media posts in your submission to Tesla’s form

• Provide your personal details in the form

• You have until July 17th, 2025, or until Tesla receives 10,000 entries, whichever comes first.

You can find the official submission form and all region-specific details on Tesla's website.

Follow

Follow

Follow