$AVGO (Broadcom) #earnings are out:

Is $ORCL actually attractive now?

A little late to the $AVGO party after their release, but here are some of my thoughts. $73 billion in AI backlog to ship over the next 18 months. Not guidance but booked orders. Q4 revenue hit $18B (+28% YoY). AI semiconductors alone did $6.5B, up 74%. Q1 guide? $8.2B in AI

Expecting more Data Center delays going forward.

$IREN $WULF $CIFR Bitcoin miners pivoting to AI datacenters had a beautiful arb: existing power + permits while everyone else waited years. The SPEED Act compresses that edge. Still like the demand story (50GW shortfall by 2030), but the margin of safety just got thinner.

The SPEED act is not bullish for $IREN $CIFR and other players.

The scale and scope of Micron’s guidance raise is similar to the “Big Bang” Nvidia quarter in May 2023 that sparked the entire AI trade I chronicled it in a chapter in my book

This is an incredible story by Reuters: they reveal that China now has a working EUV lithography machine prototype, which makes China the only country in the world to have succeeded at replicating the ASML technology, years earlier than anticipated. The machine apparently "fills

"Despite significant efforts, we are disappointed to be unable to meet demand from our customers, across all market segments." $MU

$VRT continues to impress here.

$NVDA Colette Kress said something on the NVIDIA call that I haven't seen anyone pick up on. She acknowledged that Chinese competitors, bolstered by recent IPOs, are making progress and "have the potential to disrupt the structure of the global AI industry over the long term."

Hyperscaler CapEx expectations for 2026: approaching $700B. That's up $120B since January. $NVDA's $78B Q1 guide confirms the acceleration. Follow the money downstream. $POWL $HUBB $VRT $CRDO $ALAB $ANET $CDNS The GPU quarter gets the headline. The infrastructure buildout gets

Beijing is considering a semiconductor incentive package worth up to $70 billion. Separate from the existing $50 billion Big Fund III. Xi Jinping has framed chip independence as a "whole-nation" effort. Chinese officials have actively discouraged purchases of $NVDA's

Took a break. The world didn't. War in Iran. Supreme Court killed the tariffs. Oil past $100. CPI at 3.3%. Software stocks down 30%. Meanwhile, hyperscalers committed $700 billion in capex for 2026. Every one of them says the same thing: supply-constrained, not

Software stocks down 30% in 2026. Infrastructure stocks holding up. The market is finally pricing what I've been writing about for months: the physical layer is harder to disrupt than the application layer. An AI agent can replace your SaaS tool overnight. It can't replace a

$VRT was one of our Best Ideas for 2026. Time for a check-in. The stock is up 60% year to date. Up 336% over twelve months. It joined the S&P 500 in March. These are the headline numbers. Here's what's underneath them. Vertiv reported Q4 2025 in February. Organic orders were up

$VRT reports Q1 2026 on Tuesday. The stock is up 60% year to date. Up 319% over twelve months. It joined the S&P 500 in March. Market cap now sits at $113 billion. The question is no longer whether Vertiv is executing. It's whether execution at this pace is already priced in.

Uber just blew its entire 2026 AI budget in four months. The CTO, Praveen Neppalli Naga, told The Information this week that the company is "back to the drawing board, because the budget I thought I would need is blown away already." The numbers underneath are the story. Uber

Best Ideas 2026 check-in. The ten-name picks-and-shovels basket: +23.4% YTD. $SPY: +5.5% YTD. Roughly 4x the benchmark, and Q1 earnings season hasn't even started. The lineup: $POWL +95.5% $VRT +71.5% $CRDO +17.5% $ANET +15.5% $AVGO +14.1% $HUBB +13.8% $GOOGL +7.0% $NVDA

The people who actually made money on "software" this year aren't in $PYPL or $CRM. They're in the companies selling the infrastructure the software runs on. This isn't a bottom call. It's a falling knife. The money was in the bottleneck.