A prediction published this week argues Broadcom will surpass Apple and Microsoft in market value within a decade, a call that requires the chipmaker to nearly triple while the two giants stand still. It matters because it frames the AI build-out as the largest infrastructure project of a lifetime and positions Broadcom as its central toll collector.
The numbers make the hill steep. Broadcom sits at roughly $1.76 trillion today. Apple is near $4.9 trillion and Microsoft close to $2.9 trillion. For the prediction to pay off, Broadcom must add more than $3 trillion in value while the incumbents tread water. The bull case rests on a single management projection: custom AI chip revenue exceeding $100 billion annually by 2027. Broadcom already designs the bespoke silicon that Alphabet, Meta Platforms and Anthropic rely on, and it dominates the networking fabric that stitches thousands of those chips together inside a data center.
That revenue target is the lever. If AI compute keeps compounding at its current pace, the supplier of the picks and shovels captures a growing share of the spend while Apple and Microsoft harvest mature ecosystems, iPhones and enterprise software, that grow at a more pedestrian clip. The argument is essentially that the fastest-growing slice of technology eventually rewrites the leaderboard, no matter how dominant the incumbents look today.
The counter-argument is written into the semiconductor cycle itself. Broadcom’s growth is concentrated in a handful of hyperscale customers, any of whom could decide to design more silicon in-house. A cyclical downturn in chip demand, or a cooling of AI capital expenditure, would stall the trajectory in a hurry. Apple and Microsoft, meanwhile, are cash machines with their own AI ambitions and the balance sheets to pivot aggressively. A ten-year horizon is long enough for several boom-bust cycles to play out.
The deeper point, stripped of the horse-race framing, is that the center of gravity in technology is migrating toward infrastructure. Broadcom offers one of the purest exposures to that shift, but purity cuts both ways, it lacks the diversified moats that let the giants absorb a missed cycle. Owning the runway means accepting the cyclicality that comes with it.
Ten years is a long time to hold a bet that depends on a single customer set and a single secular trend. The prediction is useful less as a price target than as a reminder that market-cap leaderboards are not permanent fixtures. Whether Broadcom actually overtakes Apple and Microsoft or not, the next decade will settle whether AI infrastructure becomes a utility-like toll road or a commoditized battlefield. The chips will keep shipping either way.