-
Broadcom (AVGO) rose 3.44% to $332.77, AI income hit $8.4B up 106%, Q2 AI steering is $10.7B. Nvidia (NVDA) reported income of $68.13B up 73%, Data Center income of $62.31B, however shares fell.
-
Broadcom’s CEO gave traders a concrete $100B AI chip income goal by 2027 with OpenAI as a buyer, offering the narrative readability Nvidia couldn’t ship regardless of sturdy outcomes.
-
The analyst who known as NVIDIA in 2010 simply named his high 10 AI shares. Get them here FREE.
CNBC’s Kristina Partsinevelos put it plainly after Broadcom (NASDAQ:AVGO) reported earnings this week: “Broadcom just did what Nvidia couldn’t explain.”
That’s the entire story. But let’s unpack why it issues.
Nvidia (NASDAQ:NVDA) reported what ought to have been a celebration. Revenue hit $68.13 billion, up 73% year-over-year. Data Center alone generated $62.31 billion. The inventory nonetheless fell. Partsinevelos nailed the rationale: “Every single person knows Nvidia now, and it takes more than a beat to just move the needle.”
READ: The analyst who known as NVIDIA in 2010 just named his top 10 AI stocks
When possession saturation is that this excessive, the bar is not “beat estimates.” The bar is “change the narrative.” Nvidia did not try this. Broadcom did.
Broadcom’s CEO delivered one thing traders not often get: a concrete quantity with a timeline. Hock Tan gave traders a transparent line of sight to greater than $100 billion in AI chip income by 2027. He additionally confirmed OpenAI as a buyer and informed the market that the software program enterprise will not be damage by AI.
That final level is underappreciated. Broadcom’s VMware-driven software program section is a recurring income machine. The bear case was all the time that AI disruption might cannibalize it. Tan killed that concern straight.
The numbers backed him up. AI income hit $8.4 billion in Q1, up 106% year-over-year, above forecast. Then Tan guided Q2 AI semiconductor income to $10.7 billion in opposition to complete Q2 income steering of roughly $22 billion, up 47% year-over-year.
Here’s the one-layer-deeper perception most protection misses. Broadcom would not compete with Nvidia head-to-head. It builds customized AI accelerators, known as XPUs, particularly designed for particular person hyperscalers. Think of it this manner: Nvidia sells the identical GPU to everybody, like a ironmongery store promoting hammers. Broadcom builds a customized software for every contractor’s actual job. Google, Meta, and now OpenAI don’t need off-the-shelf silicon once they’re working trillion-parameter fashions at scale. They need purpose-built chips that maximize effectivity for his or her particular workloads. That’s Broadcom’s moat.