$Vicor(VICR)$ $Everpure(P)$ $Astera Labs, Inc.(ALAB)$ $Marvell Technology(MRVL)$ Just a heads up. Vicor booked over $100 million in orders during the first two weeks of October alone. For perspective, their total physical product revenue for the entire third quarter was $130.5 million. Imagine panic-selling the recent dip when VICR dropped a solid 68.3% gross margin and $66.6M in high-margin royalty revenue. The bears seem completely blind to the model shift. This isn't a components factory, it's an IP licensing machine collecting pure-profit rent from the AI hyperscalers. With a $406 million backlog and
$Marvell Technology(MRVL)$ The price action here is pretty weak and predictable. It pops just to get sold off shortly after, and I'm saying that as someone who's bullish.
$Marvell Technology(MRVL)$ I ended up selling a cash-secured put on this one. Couldn't help myself. If it gets assigned, I just get a really solid entry price.
$Marvell Technology(MRVL)$ The AI trade is maturing and getting more scrutiny. With each rolling correction, more of the names that ran up are going to fade. This one looks like it can stick around through the late innings.
$Marvell Technology(MRVL)$ On the bright side, it is $150 from the 52-week low. That is a solid return for longs. Jensen seems to know the market and has been projecting market caps with some accuracy.
$Vicor(VICR)$ $NVIDIA(NVDA)$ $Astera Labs, Inc.(ALAB)$ $Marvell Technology(MRVL)$ It's a good thing Vicor Corp only sells 20-25% of its sales into data centers. Imagine if that were 100%. Revenue breakdown: • Industrial & Test (40%-45%): Factory automation, ATE chip testing, and medical imaging. • Aerospace & Defense (25%-30%): Rugged military computing, satellites, and radar. • Data Center & AI (20%-25%): Scalable power modules and high-density computing platforms. • Automotive/EVs (5%-10%).
$Microsoft(MSFT)$ NOV20 550C still looks actionable to me. Zone: $529.22 to $536.68 Invalidation: close under $525.02 Objectives: $543.81, then $553.72 The math: up 2.0% to the first objective and 3.9% to the second, against 1.5% risk on the invalidation, about 2.6 to 1 to the second. Microsoft closed at $532.95 on October 21 after buyers held the $525.02 low for a second day in a row. In plain terms, sellers pushed and the floor did not move. Earnings land on October 28, and that report will test whether cloud growth can carry the story, so holding above that low into the print keeps room open toward the October highs. Watching $543.81 first, the October 16 high. A close above it puts the high near $553.72 in
$Meta Platforms, Inc.(META)$ Meta popped 5% after hours to $733 despite a headline EPS miss ($3.83 vs $5.89 est). The miss was entirely due to a ~$10B settlement charge with U.S. State AGs. Without that one-time hit, normalized EPS was around $7.40, easily beating consensus by a wide margin. Key highlights: Revenue: $64.6B vs $63.4B estimate (Ad revenue +26% YoY) Q4 Revenue Guidance: $75B vs $74B estimate FY26 CapEx: Raised to $140B vs $138B estimate This is a meaningful signal for the AI ecosystem. Even with major legal expenses pushing quarterly free cash flow negative (-$2.1B), Meta raised CapEx to $140B. They are not stepping back on compute buildout. That sustained commitment is driving an after-hours lift across AI tech and semi names,