Peter Rodriguez
11.3K posts
These guys are trying to prove that Tesla FSD is better than Waymo and their car attempts to murder them no less than 3 separate times lmfao You gotta see the ending too
$AAOI is my biggest position. Let me show you exactly why using their own words. On the Q4 earnings call, CEO Thompson Lin was asked about the $1B+ revenue guide for 2026. His response: "The demand is much, much bigger than $1 billion. That's the number we feel minimum 99% confident we can deliver." Then the CFO said this about the $378M monthly run rate target for mid-2027: "This revenue level is limited by our production capacity and supply chain — not market demand, which we believe is much larger." Read that again. The ceiling on this company is not customers. Not competition. Not pricing. Purely how fast they can build fabs. Now here's what nobody is modeling correctly. That $378M monthly number — $4.5B annualized — is not the ceiling. Management explicitly said customer demand exceeds even that figure. On capacity, here's the number that floored me: End of 2025: 90,000 units/month of 800G capacity. End of 2026 target: 500,000 units/month. 5.5x capacity increase in 12 months. Here's how that ramp actually plays out: - 800G firmware completes: mid-March - 800G volume ramp begins: Q2 2026 - 500,000 units/month online: end of 2026 - $378M monthly run rate: mid-2027 And here's the part the market is completely missing on capex efficiency: 800G and 1.6T are manufactured on the exact same production line. Every dollar $AAOI spends building 800G capacity is automatically 1.6T capacity. Competitors building separate lines are burning capital twice. $AAOI is not. Then on hard orders: "Within less than three months — $100M+ in 800G orders. $200M+ in 1.6T orders." From a company doing $134M in total quarterly revenue today. "It's not a demand issue" — said three separate times on the call. The question for $AAOI is not whether demand exists. That's been answered. The question is purely execution. And they just committed $300M to triple their laser manufacturing capacity in Texas to answer it. Full breakdown + DCF on my Substack(Link in Bio + Below)
$SNDK made more profit in one quarter than it made across the prior three years combined. Agentic AI is driving demand for NAND-backed storage across inference workloads, with SanDisk’s high-capacity enterprise SSDs positioned as a key part of the stack.
wish me luck
This is one of the more insane articles I’ve read in awhile
