Brian Kurtz

Groq rack is 8 LPU, 1 FPGA, 1 CPU, 1 BF DPU per tray and 32 trays in a rack.

$ORCL CEO WHY DC DON’T NEED TO BE NEAR POPULATION CENTERS, YET: “Inferencing is very rapidly growing everywhere and anywhere... it’s because of higher and higher utilization of the models themselves and also new use cases... inferencing is going to have a huge amount of demand.” “If what you’re doing is asking a question for your business, it’s going to take an AI model several seconds to think about it... an extra 40 milliseconds of latency from New York to Wyoming is not going to hurt you.” “The latency problem right now is not actually the location of the hardware. It’s the type of hardware that’s being deployed, and that’s why you’re seeing so much innovation around these AI accelerators.” “If you look at what Groq does, or Cerebras, or Positron, all of these different types of companies are saying not only how do we reduce the cost of inferencing, but also how can we significantly reduce the latency of it.” “That makes it much more flexible for us to put data centers where power is abundant, land is plentiful, and we can optimize for what’s available to meet this ever-increasing demand.”

I plan on owning my own Tesla Robotaxi fleet one day. And the more I run the numbers, the more I realize this new business could become one of the most powerful income opportunities I've ever seen. This is how I'm thinking about it. Based on many analyst models and Tesla’s long-term vision, a reasonable base case assumption is about ~$30,000 per year in net profit per Robotaxi to the owner. This is after things like Tesla’s platform fee, charging, tires, maintenance, insurance, and cleaning. Of course, the network is still early and Tesla is just beginning to roll this out in pilot programs in a few cities, so there’s no official real-world owner earnings yet... but using reasonable assumptions around utilization, pricing per mile, and operating costs, the math starts to get really interesting. If one Robotaxi can earn around $30,000 per year, here’s what a fleet might look like: • $100,000 per year → about 4 Robotaxis • $500,000 per year → about 17 Robotaxis • $1,000,000 per year → about 34 Robotaxis It may sound a bit crazy at first, but when you break it down, it starts to make more sense. These vehicles could potentially drive 50,000 to 100,000+ miles per year in high demand areas. If the economics land somewhere around $0.25-$0.50 profit per mile after all costs, you end up right around that ~$30k per vehicle per year range. And remember, the Tesla’s Robotaxi network is going to work a lot like Airbnb for cars. You add your vehicle to the network, Tesla handles the software, routing, payments, and rider experience, and they take a platform fee (often modeled around 25-35%). The owner keeps the rest after operating costs. Another thing that makes this interesting is the expected cost of the vehicles themselves. Tesla has talked about the purpose-built Cybercabs costing roughly $25k-$30k and Elon told me production is starting in 1 month! If that’s even close to reality, a fleet capable of generating around $1 million per year could theoretically cost somewhere around $850k-$1M in vehicles. That ROI is pretty freakin good! Now to be clear, none of this is guaranteed. I'm just thinking out loud and sharing it with you... a lot still depends on regulations, how fast unsupervised FSD scales, demand in each city, insurance costs, and how Tesla structures the network. But if the system works the way Elon has described it for years, owning a Robotaxi fleet could become one of the most powerful forms of passive income I've ever seen. And I plan on sharing the numbers with everyone on 𝕏 when the day comes. Personally, that’s why I’m paying such close attention. Bc one day, owning a fleet of autonomous Teslas working for me 24/7 might be the modern version of owning a rental property, except instead of tenants, you’ve got robots driving people around all day while you sleep. This next book of Tesla is going to be so exciting!

$CRDO Q3 call revealed two things the market is sleeping on: Weaver gearbox: 10x memory IO density. Positron building a 2TB inference XPU on it for speed! Lasers are out: ZeroFlap optics 1000x more reliable, half the power. Production ramp Q1 FY27. Listen Credo: