An a former $INTC Intel engineer I wrote my thoughts on the company back in December, 2024. I advocated investors against Intel. Have my views changed? I never owned Intel. Not when the stock ran. I worked there. I knew the failures were never about demand. They were about management and execution. Lately I bought a small number of 2028 calls. I talked to former colleagues. Under Lip-Bu Tan, the company was focused on profits. The old, inefficient guard was being cleared out. The fab was doing better. 18A yields were improving. I've always been skeptical about yields. Not just at Intel, but at any US fab, given how much talent we've lost over the years. That skepticism is starting to ease. Yesterday's call was rough. Supply is the ceiling. Substrates are short. A tragic accident shut Fab 28. The stock fell 21%. After a lot of angst, I'm holding. Here's why. Intel is two very different businesses. One designs CPUs for servers and PCs. The other is a fab that drains capex. Demand isn't the problem for either one. Execution is. Intel still shows wear and tear. But it has a better CEO than it has had in a decade, and the US government owns a stake. It's more able to fix its mistakes today than at any point I've watched.Demand is robust. And there's the bigger picture. Intel is the only leading-edge US fab, and the only one showing real signs of change. China is ahead in many ways. Depending on Taiwan is not a long-term plan. Trillions of dollars of US tech run on chips made on an island Beijing claims as part of One China. Intel can't fail. So I'll give it a chance. Call me foolish.

$GOOGL report ready! Alphabet dropped about 10% after hours. The headline said EPS of $2.12 against a $3.00 estimate. Most of that gap was an $11 billion mark-to-market loss on Alphabet's SpaceX stake. Strip that out and the quarter still missed. Just by less. Here's how I read the print,and why the supply chain matters more than the stock tonight. The real problem isn't SpaceX. It's the bill for the AI build. Capex hit $51 billion in a single quarter. Free cash flow went negative again. Buybacks went to zero. The one thing paying for it: Cloud, up 103%.

$GOOG capex for full year reiterated. $MSFT likely does the same even though both q3 #s# were soft. Prob just waiting for Rubin and turning down more blackwell shipments in q3. $MRVL buying $LITE rumor-wow. Been nibbling LITE below 900. If MRVL really pays up over 100b for LITE, $AAOI CAN go to 200 within a quarter or two IMO. Shows just how valuable having laser capacity is. $CBRS hit piece-gift if msft capex reitertated(likely)-VICR would've fell hard if this hit piece was real-it held up like the champ it is. Own both, but more VICR, way more VICR, like 30x more. Fun day.

I don’t think any wearable in market right now can accurately measure quality of sleep. So why would OURA ring have a valuation of $16B? Recently filed S-1 suggests they plan to raise $3B from IPO. $16B for selling a device that consumers trust to measure inaccurate data or a competitor figures how to make a more accurate device? This IPO makes little sense to me. Wearables can’t register those electrical signals. Instead, optical sensors track blood flow and micromovements. If you lie dead still and your pulse plummets, the algorithm might assume you’re in deep sleep, a reasonable deduction, but still a guess. - WSJ