So I've been watching the recent pullback in tech stocks, and honestly, there's some interesting opportunities forming if you're willing to think long-term. The market got spooked by geopolitical noise earlier this week, but the fundamentals underneath are actually pretty solid. Earnings and interest rates are still supporting things, and if you look at what's really moving the needle, the AI capex story is far from over.

Taiwan Semi already guided 2026 capex to $52-56 billion back in January—way above 2025's $40.9 billion. And the hyperscalers? They're projected to spend roughly $530 billion in capex this year, up from around $400 billion last year. That's not slowing down. Meanwhile, Q1 2026 tech sector earnings estimates have jumped to 24% growth from just 18% a month ago. This is the kind of backdrop where buying dips actually makes sense.

I've been looking at ServiceNow (NOW) pretty closely. It's gotten crushed—down nearly 50% from its January highs—but that's actually interesting because the company itself is executing well. They're not some software dinosaur getting disrupted by AI; they're actively integrating it into everything they do. Deepened partnership with OpenAI, expanding Claude integration with Anthropic, building what they call an "AI control tower for business reinvention." The Q4 numbers showed 244 transactions over $1 million in net new ACV, up 40% year-over-year. Revenue hit $13.28 billion in 2025, more than double 2021. GAAP earnings grew 22% to $1.67 per share.

Wall Street's projecting 20% revenue growth for 2026 and 18% for 2027, with adjusted earnings climbing 18% and 20% respectively. CEO Bill McDermott just bought $3 million of shares himself, calling it the best entry point in a while. The stock's average price target suggests about 70% upside from here. If it returns to those January highs, you're looking at close to 100% upside. That's the kind of risk-reward that makes sense for patient investors.

Now, Celestica (CLS) is a different animal—more of a pick-and-shovel play in the AI infrastructure space. They design and build the servers, switches, and data center hardware that the hyperscalers actually need. Revenue grew 29% in 2025 to $12.39 billion, and they've more than doubled revenue since 2021. Adjusted earnings jumped 56% last year, GAAP EPS over 90%. The company just guided for sustained momentum into 2027, with 37% projected revenue growth for 2026.

CLS is down about 25% from its November highs right now, and it's trading at 30x forward earnings—50% below its recent peaks. That's a pretty compelling entry for a stock that's been crushing it. The guidance suggests $23.66 billion in revenue by 2027, with adjusted earnings expanding 46% and 43% in the next two years. Analysts have 15 out of 18 recommendations as Strong Buy, and the Zacks Rank just upgraded to #2 Buy.

The broader point here is that these tech stocks are getting hit in the short term, but the structural demand for AI infrastructure and enterprise AI solutions isn't going anywhere. If you've got a longer time horizon and can tolerate some volatility, these kinds of dips are exactly when you should be thinking about accumulating quality. The earnings growth is there, the capex cycle is accelerating, and valuations have compressed enough to offer real margin of safety. That's the setup I'm watching right now.