So we're halfway through 2026 now and I've been thinking back to what looked promising at the start of the year. The AI infrastructure play is still the dominant narrative, and honestly, most of the big chip and cloud names have delivered on expectations. Let me walk through the ones that actually caught my attention when I was building my watch list.

Obviously Nvidia was the safe bet everyone was talking about. The data center spending story just keeps getting bigger - they're projecting $3-4 trillion in global capex by 2030, which is wild. As long as hyperscalers keep throwing money at AI infrastructure, Nvidia's probably going to keep winning. It's been the face of this whole buildout since 2023 and I don't see that changing anytime soon.

But here's where it got interesting for me. AMD has been quietly closing the gap in the GPU space. Their data center revenue was supposed to hit that 60% compound annual growth rate they were targeting, and they were only at 22% growth back in Q3. If they actually pulled that off, that's the kind of acceleration that gets people excited. Same story with Broadcom - they went a different route with custom AI accelerators instead of competing head-to-head on GPUs. Their AI semiconductor revenue jumped 74% year over year and they were expecting over 100% growth in Q1 of their fiscal 2026. That kind of momentum catches people's attention.

Taiwan Semiconductor is the play that makes sense if you want to stay neutral in all this. They're the foundry everyone uses - Nvidia, AMD, Broadcom all outsource to them. So as long as AI spending keeps climbing, TSMC wins regardless of who's winning the chip wars.

On the mega-cap side, Alphabet finally got respect for their AI game. Gemini went from being written off as inferior to actually being competitive with the best stuff out there. Plus they've got Google Search as a cash machine and Google Cloud as a real business now. It's hard to find a weak point.

Meta was interesting because the market punished them for their capex plans, but people weren't paying attention to how strong the actual business was performing. Their revenue was up 26% year over year in Q3, driven by AI effects on their platforms. That discount that opened up looked like a genuine buying opportunity to me.

Amazon only moved about 3% for most of 2025, which seemed like people were sleeping on the business. Their cloud and advertising divisions were firing on all cylinders. Both of those segments are built to thrive in an AI-heavy world, so the stock had room to run.

PayPal got absolutely crushed - down 30% - but the business wasn't nearly as bad as the stock price suggested. They were actually delivering solid earnings per share growth through 2025 and running share buybacks. At 11.5 times forward earnings, it looked cheap.

Then there's The Trade Desk, which had a brutal year down 70%. They messed up their migration to their AI platform Kokai and lost some clients to Amazon because of it. But Wall Street was still projecting 16% revenue growth for 2026 at a 20 forward P/E ratio. That setup looked like a potential turnaround candidate.

MercadoLibre rounded out the list as the Latin American e-commerce and fintech story. They were up 20% for the year but pulled back over 20% from their July highs. They've been the dominant player in their region and every pullback over the years has turned into a buying opportunity eventually.

Looking back now at mid-2026, the stock advisor top 10 type picks that actually focused on the AI infrastructure theme and the companies executing on their growth plans have mostly held up. The ones that had disappointed performances in 2025 - like Meta and Amazon - did bounce back like expected. The key was having a real stock advisor top 10 framework going into the year, not just chasing whatever was hot at the moment.