TSMC: The $2 Trillion Underdog Set to Outpace Broadcom and Nvidia by 2028

Oh, diary. Another day spent staring at stock charts like they’re the plot of *Bridgerton*. Will Broadcom crack the $2 trillion ceiling? Or will TSMC swoop in like a well-timed witty retort at a dinner party? Let’s just say I’ve been thinking about this more than I’ve thought about anything else this week (and that includes whether or not I should invest in a bread maker).

Units of Bread Makers Considered: 1. Hours Spent Researching AI Semiconductors: 47. Number of Times I Pretended to Understand Advanced Process Nodes: At least 12.

Here’s what we know: Big tech companies are set to pour $375 billion into artificial intelligence infrastructure this year, according to UBS analysts. Next year? That number balloons to $500 billion. And where does all that money go, you ask? Straight to semiconductors-the unsung heroes of our AI-driven future. But here’s the rub: while everyone’s fawning over Nvidia (NVDA), there’s another player quietly plotting its ascent to the $2 trillion club. Spoiler alert: it’s not Broadcom.

Broadcom: A $2 Trillion Contender?

Now, let me preface this with a confession: I once thought “Broadcom” was some sort of obscure yoga pose. Turns out, it’s a massive company with fingers in both hardware and software pies. Their AI chip business is particularly spry, growing revenues by 46% last quarter alone, reaching a cool $4.4 billion. Management expects even more growth-$5.1 billion in AI semiconductor revenue this quarter, with growth accelerating to roughly 60%.

But wait, there’s more! (Isn’t there always?) They’ve also got VMware under their belt now, which has proven to be quite the cash cow. Over 87% of VMware customers have transitioned to the subscription model, fueling double-digit growth in annual recurring revenue. Impressive, no? And yet…

Broadcom shares are trading at a forward P/E ratio of 45. Yes, you read that right. For a company whose top-line growth hovers around 20%, that feels a bit like paying for a Michelin-starred meal but only getting avocado toast. Sure, there’s margin improvement from VMware and strong AI sales, but can it really justify such a lofty valuation? I’m skeptical. It’s like dating someone who looks great on paper but insists on quoting Nietzsche during brunch.

Loading widget...

TSMC: The Quiet Giant

Enter Taiwan Semiconductor Manufacturing Company (TSM), the behind-the-scenes genius making all those shiny GPUs possible. Without TSMC, neither Nvidia nor Broadcom would have chips powerful enough to train AI models or run inference tasks. In fact, TSMC controls over two-thirds of the semiconductor manufacturing market. That’s not dominance; that’s an empire.

Here’s why I find TSMC so fascinating: they operate in a virtuous cycle. They win big contracts because they’re ahead technologically, then use that revenue to invest in even better technology. Case in point: their latest process node, N2, reportedly charges a whopping 66% premium per silicon wafer compared to the previous generation. Companies are lining up to pay it, desperate for that extra bump in power efficiency. Meanwhile, TSMC maintains its margins despite the usual teething problems associated with new nodes. Clever, isn’t it?

Management predicts AI-related revenue will grow by mid-40% annually from 2024 through 2029. While AI chips remain a small slice of TSMC’s overall pie, this growth translates to about 20% total revenue expansion. Operating earnings? Likely to exceed that figure, thanks to their ability to keep gross margins stable as they roll out newer processes.

Loading widget...

And here’s the kicker: TSMC trades at a much more reasonable earnings multiple of 24. If they maintain this multiple while growing earnings at 20% annually, reaching $2 trillion by 2028 seems entirely plausible. Unlike Broadcom, which feels like it’s already priced for perfection, TSMC strikes me as the smarter bet. It’s like choosing between a flashy sports car and a reliable sedan-you know which one will still be running five years later.

In conclusion, dear diary, I think I’ll put my hypothetical millions on TSMC. Not because I’m brave or particularly insightful, but because sometimes life rewards patience and pragmatism over flashiness. Also, because I’ve decided against buying a bread maker. Units of Bread Makers Purchased: 0. 🍞

Read More

2025-08-30 19:35