Praktisch Gesprochen

Praktisch gesprochen über Finanzen, Tesla, SpaceX & langfristigem Denken.
Practically speaking money, Tesla, SpaceX & long-term thinking.

NOT FINANCIAL ADVICE

[Praktisch Gesprochen @PraktischPS]

Part 3 is about KPIs that matter to compute buildout. Drops here soon.

[Praktisch Gesprochen @PraktischPS]

The 22-hour coast for return-to-launch-site catch is a great proof of Starships orbital capability. If both stages come home, cadence and cost per V3 satellite drop sharply. Practical filter: first-attempt ship-catch reliability will set near-term deployment pace more than factory output.

[Praktisch Gesprochen @PraktischPS]

@imenzo The shift is clear—compute infrastructure at Starbase now exceeds prior Starship capex. Long-term advantage is integrated power + launch + satellite demand. How does Arizona fit in?

[Praktisch Gesprochen @PraktischPS]

Live unsupervised @robotaxi service is currently in ~8–9 U.S. cities/regions (e.g., multiple Texas cities plus Florida markets). Expansion has been rapid in 2026. Reaching 25 cities could occur in ~6–12 months at the recent pace. $TSLA

[Praktisch Gesprochen @PraktischPS]

This move happened quickly and weather in the future the allegations proof to be true or not remains to be seen. This is to separate the person from the company. Earnings call is going to be interesting. Fundamentals have not changed. Nvidia is still the chip supplier it was a week ago $NVDA

[Praktisch Gesprochen @PraktischPS]

@JoeTegtmeyer @aelluswamy It’s a key unlock for robotaxi. I’m sure scaling will accelerate from here on out

[Praktisch Gesprochen @PraktischPS]

A separate group of large holders (~7.8B shares total, including Musk) has a longer staggered lock-up:
• Staged releases starting after Q4 2026 results (20%), then day 280 (Mar 18, 2027), after Q1 2027, day 340 (May 17, 2027), etc.
• Elon Musk’s shares (~6.4B) have a 366-day lock-up with no early release — first tradable around June 14, 2027 (day 366 falls on a Saturday, June 12).

[Praktisch Gesprochen @PraktischPS]

Live unsupervised @robotaxi service is currently in ~8–9 U.S. cities/regions (e.g., multiple Texas cities plus Florida markets). Expansion has been rapid in 2026. Reaching 25 cities could occur in ~6–12 months at the recent pace. $TSLA

[Praktisch Gesprochen @PraktischPS]

So so this is how it’s going to be from here on out. We also will see faster scaling from here on out. Don’t miss the signs, they are the opportunity. Bullish

[Robotaxi @robotaxi]

Robotaxi now runs 24/7

Every city we're available in

[Praktisch Gesprochen @PraktischPS]

@MooreDividends Ja ich schon. More of the same ist ja auch langweilig. Longterm Investor Info aus praktischem Blickwinkel gibts bei mir. Freu mich wenn’s wem gefällt. 🙋🏻‍♂️

[Praktisch Gesprochen @PraktischPS]

7/7 - Investor orientation:
The companies that supply or control the slowest layers (power equipment, advanced packaging, high-bandwidth interconnect, liquid cooling) have more durable positioning than pure GPU buyers. The stack is the constraint.

[Praktisch Gesprochen @PraktischPS]

The Compute Stack — Layer by Layer 🧵
Buying GPUs is the easy part.
Turning them into working frontier capacity requires the full physical stack to arrive at the same time.
Here’s what compute actually takes. (read more...)

[Praktisch Gesprochen @PraktischPS]

6/7 - Cooling
Rack densities are rising (120+ kW racks). Liquid cooling and CDUs are becoming standard. Thermal limits now constrain density. $SCHN.NE $VRT $ETN

[Praktisch Gesprochen @PraktischPS]

@brewmarkets If in doubt zoom out. Works every time. If you care for longterm practical investment takes, I am interested in getting to know you.

[Praktisch Gesprochen @PraktischPS]

AWS scaling GPUs every 14 seconds on the ground is real, but so are the permitting, interconnection, and power-queue constraints. Orbital proposals (including SpaceX’s large filings) are still early; the durable advantage appears only if launch cost and thermal management improve enough that continuous solar beats the terrestrial bottlenecks. Timeline gap is large today—the crossover is a cost-curve question, not a 2027 certainty. Which constraint do you see relaxing first?

[Praktisch Gesprochen @PraktischPS]

Dual catch on Flight 15 is the first real test of whether the tower becomes a high-utilization asset rather than a one-stage recovery tool. If both stages return intact, the marginal cost of the next launch drops sharply because the expensive hardware is already back. Even a partial success compounds the reusability curve faster than another expendable flight. What metric from this attempt would you watch most closely for the cost trajectory?