G.L.A. Invest Research x Coach Kyle — Tesla, SpaceX, Palantir & The AI Infrastructure Stack
A G.L.A. INVEST RESEARCH publication, co-authored with Coach Kyle
Happy Tuesday to the Greatest Living Americans!— Coach Kyle here with the G.L.A. INVEST RESEARCH team. Our mission hasn’t changed since day one: keep you “in the know.” with what’s really going on with the stock market and its best investments.
Retail investors constantly get flooded with noise from Wall Street voices who are often paid, incentivized, or fee-motivated to tell you what sells their business model, not what’s in the best interest of most retail investors. I built G.L.A. to be different — independent insights and perspectives of an Ex Broker who’s unafraid to call things as I see them, so investors of every economic background have a real shot at building generational wealth. This week we’re diving deep into my favorite topic, the Elon Musk ecosystem (Tesla, SpaceX, Terafab) and the broader AI infrastructure names our members keep asking about: Palantir, Nvidia, Cerebras, Micron, and SanDisk.
Tesla FSD: The Safety Story Nobody’s Telling Right
Here’s something we think most retail investors — and honestly most financial media — completely misunderstand: Tesla’s Full Self-Driving system is now producing crash-rate data that, if it holds up, is genuinely remarkable. Tesla’s own released safety data claims FSD (Supervised) vehicles see roughly 7x fewer major collisions and 5x fewer off-highway collisions compared to a human driving the same car manually. Independent trackers have cited figures suggesting Tesla vehicles running FSD are seeing about one crash per 5 million miles, versus roughly one crash per 700,000 miles for the average U.S. driver — something like a 700% safety improvement over typical human driving.
Humans for the most part suck at driving, and now that every driver is holding an iPhone the risks are even higher. If you don’t believe me, just take a drive on the Southern State Parkway in Long Island and you will notice every 2 miles or so you will see flowers and a make shift Shrine where someone died in a car crash, not to mention the number 1 cause of death for people under 35 is by car crash. The crazy thing about Tesla FSD is that it will continue to get safer and safer until the point where it will become illegal for humans to drive. I know that a lot of us love to drive and would hate to see our driving rights taken away, but if FSD can get to be 1000 times safer and we can end most driving deaths I think most will approve of it. My advice is if you havnt’t tried Tesla FSD you should do so ASAP because it will blow you away!
Robotaxi & Cybercab: Where the Revenue Comes From
Tesla’s Cybercab — the two-seat, no-steering-wheel autonomous vehicle — began pilot production in February 2026, with Elon Musk confirming a real production ramp starting in April 2026., with a cost of under 30K. Tesla has filed for permits to run up to 5,000 robotaxis in Nevada’s Clark County (Las Vegas/Henderson) within the first 12 months of launch, and the company’s own roadmap targets robotaxi expansion across Dallas, Houston, Phoenix, Miami, Orlando, Tampa, and Las Vegas through the first half of 2026. The bull case here is straightforward: every mile driven without a human behind the wheel converts a depreciating asset (a parked car) into a revenue-generating one, and Tesla is betting the FSD safety data above is the unlock that gets regulators and riders comfortable at scale.
Terafab: Tesla + SpaceX’s Trillion-Dollar Chip Bet
This is the story most of Wall Street hasn’t fully connected yet. Terafab is a jointly developed semiconductor fabrication plant between Tesla, SpaceX, and xAI/Intel, announced by Musk in March 2026, designed to produce over one terawatt of AI compute capacity annually. SpaceX has filed for an initial $55 billion investment with total build-out costs potentially reaching $119 billion across all phases, and as of early August 2026, SpaceX and Tesla committed an initial $16.8 billion toward construction in Grimes County, Texas. The facility is expected to span roughly 100 million square feet — about 10 times the size of Tesla’s Gigafactory Texas — and Musk has framed the motivation bluntly: he’s worried existing chip suppliers like TSMC, Samsung, and Micron won’t be able to keep up with Tesla, SpaceX, and xAI’s combined AI chip demand within three to four years. This is a direct vertical-integration play, and when it works, it will remove a major bottleneck across the entire Musk ecosystem — robots, robotaxis, Starlink, and orbital compute all draw from the same chip pipeline.
SpaceX: From Rockets to a Trillion-Dollar AI Cloud Business
SpaceX’s first earnings report since its June 2026 IPO turned heads across the market. Q2 2026 revenue hit $7.81 billion, up 92% year-over-year, with adjusted EBITDA of $3.54 billion (up 191%) and a net loss that shrank to $541 million from roughly $1 billion a year earlier. Management guided toward a $100 billion annualized revenue run rate by the end of 2026, and Musk pulled forward SpaceX’s trillion-dollar annual revenue target from 2031 to 2030, floating a “non-zero chance” it lands as early as 2029.
The segment breakdown is the real story: Launch brought in $962 million, Starlink did $4.29 billion (up 66%, with subscribers doubling to 12 million), and — most strikingly — the AI/cloud compute business did $2.56 billion, up roughly 247% year-over-year, driven by contracted cloud deals with Google, Anthropic, and Reflection AI worth $14.1 billion. That AI segment now represents about a third of total SpaceX revenue and came from essentially nothing about a year ago. SpaceX finished the quarter at 1.4 gigawatts of compute capacity, targeting 2 GW by year-end and roughly 10 GW by the end of 2027, with a formal Nvidia partnership (Vera Rubin architecture) underpinning a planned orbital AI compute program called Star Mind. The bear case investors should weigh: capex hit $18 billion in the quarter (roughly 6x a year prior), with $15.8 billion going directly into AI infrastructure, and the AI segment itself lost $1.26 billion at the operating level — this is a high-spend, high-conviction bet on sustained compute demand.
COACH KYLE’s 2 CENTS: SPACEX is a longterm story and a longterm investment. Volatility will continue but that’s ok, because G.L.A INVE$T, Coach Kyle, and our readers fully understand that volatility is simply the price of admission for exponential growth and compounding interest. We had extreme volatility with AMZN, we had it with PLTR and TSLA, we also had it with NVDA and every other growth stock that 10 X’d or more over the last 15 years. Our volatility tolerance is battle tested and proven as we move through the early stages of the greatest innovation boom since the early 1900’s.
Palantir: The Other Side of the AI Stack
Where SpaceX is building physical infrastructure, Palantir is building the software control layer sitting on top of it — and Q2 2026 numbers were, in Alex Karp’s own words, “otherworldly.” Total revenue hit $1.935 billion, up roughly 93% year-over-year, beating consensus of about $1.8 billion for a 12th consecutive quarter of accelerating growth. U.S. revenue reached $1.573 billion (up 115%), U.S. commercial revenue hit $764 million(up 149%), and U.S. government revenue grew 90% to $809 million. Palantir posted adjusted free cash flow of $1.22 billion (a 63% margin), GAAP net income above $1 billion, and a “Rule of 40” score of 155%. Management raised full-year 2026 revenue guidance to a range of $8.15–$8.16 billion, implying roughly 82% annual growth. Net dollar retention sits at 157%, remaining deal value is $13.1 billion (up 83%), and the company closed 220 deals worth over $1 million, 73% of which exceeded $10 million. Karp’s core thesis — that companies handing raw data to frontier AI labs risk losing their competitive edge to those same labs — is resonating with enterprise buyers who want to retain control of their own data and models.
Nvidia, Micron, SanDisk & Cerebras: The Picks-and-Shovels Layer
Rounding out the AI infrastructure trade, several “picks and shovels” names are continuing to impress.
Company (Ticker) | Recent Signal | Source |
|---|---|---|
Nvidia (NVDA) | Analysts expect Q2 FY2026 revenue of ~$94–95 billion, above the $91B guide, on Vera Rubin demand and rising cloud capex | Investing.com, Reddit/analyst tracking |
Micron (MU) | Fiscal Q3 2026 revenue of $41.46 billion, up 74% sequentially, driven by sold-out HBM/AI memory demand | Tickeron |
SanDisk (SNDK) | Fiscal Q3 2026 revenue rose 251% YoY to $5.95 billion on AI data-center storage demand; some analysts set price targets as high as $3,250 | Yahoo Finance, YouTube analyst coverage |
Cerebras Systems (CBRS): Still Early In Its Ramp
One of the most-discussed upstarts across our network is Cerebras Systems, and this week’s action shows exactly why. Cerebras stock surged 15.1% to close at $251.98 on Monday ahead of its flagship Supernova event in San Francisco, with CEO Andrew Feldman telling Investor’s Business Daily the company is “still early in its ramp of production of processors and infrastructure for artificial intelligence” and has a long growth runway ahead.
The speed pitch is the core of the bull case:Cerebras’ Wafer-Scale Engine keeps AI model weights on-chip using 44 gigabytes of SRAM per wafer-size chip, avoiding the memory-bandwidth bottleneck that constrains traditional GPU-based inference. That architecture is now powering OpenAI’s new “Ultrafast mode” tier for GPT-5.6 Sol, running up to 750 output tokens per second — up to 14 times faster than standard processing. As Feldman put it: “There’s no market for slow search. There’s no market for dial-up. There’ll be no market for slow AI.”
On the numbers, Cerebras’ core revenue jumped 103% in Q2 2026, and despite beating non-GAAP expectations, the stock initially dropped last week before this week’s rally — a reminder that even strong prints can disappoint if expectations run ahead of results. The company guided full-year 2026 core revenue to $885 million, up 74% from 2025, and forecast revenue to more than triple in 2027, backed by a $25.4 billion remaining performance obligations (RPO) backlog. “AI demand is accelerating,” Feldman said. “Right now the market is growing so fast, everyone’s just trying to keep up.”
On supply chain, Cerebras has secured additional wafer capacity from TSMC (TSM) and added new production lines at Flex (FLEX), Sanmina (SANM), and Rocket EMS, with major customers including OpenAI and Amazon Web Services (AMZN). Notably, Cerebras avoids the current AI chip bottlenecks entirely — it doesn’t use HBM, CoWoS packaging, or 3-nanometer fabrication, instead running on TSMC’s more available 5-nanometer node for at least another 18 months, which gives it a real supply-side edge over Nvidia in the near term. UBS analyst Timothy Arcuri reiterated his buy rating and raised his price target from $320 to $330 ahead of the Supernova event, where Cerebras is expected to unveil its next-gen CS-4 system, with CS-5 targeted for the second half of 2027 and a roadmap aiming to double inference speed each year.
Cerebras trades at a $72.65 billion market cap with a 52-week range of $160.81 to $386.34 and a beta of 3.26 — this is a high-volatility, high-conviction name, not for investors who have a low volatility tolerance.
What This Means For G.L.A NATION:
The common demoniator across Tesla, SpaceX, Palantir, Nvidia, Micron, SanDisk, and Cerebras is that the AI buildout is no longer just a chip story — it’s a full-stack story spanning launch vehicles, satellites, orbital compute, memory, storage, and the software layer that sits on top of all of it.
Coach Kyle’s read: this is exactly the kind of multi-year structural shift that rewards patient, research-driven investors over those chasing headlines — but it also comes with real volatility, with real up’s and down’s. Knowing what we own is imperative, watching the company not the stock is imperative, sticking with the strategy and our core principles is what we will continue to do! Until next time my G.L.A’s Peace out!!
Disclaimer: This newsletter is published by G.L.A. INVEST RESEARCH for informational and educational purposes only. Nothing in this issue constitutes financial, investment, legal, or tax advice, and it should not be relied upon as a recommendation to buy, sell, or hold any security. Coach Kyle is not a licensed financial advisor, registered investment advisor, or broker-dealer. All investing involves risk, including the potential loss of principal. Past performance and projected figures (including company-issued guidance) are not guarantees of future results. Please consult a qualified, licensed financial professional before making any investment decisions. G.L.A. INVEST RESEARCH’s mission is to help keep independent investors “in the know” with balanced, experienced perspectives — free from fee-driven or product-motivated bias — but all final investment decisions and their outcomes are solely your own responsibility.