Nvidia has become the face of the AI boom, but investors are now looking for the next semiconductor winner that can turn artificial intelligence demand into the kind of long-term growth Nvidia enjoyed. Broadcom, Marvell, AMD and Intel are among the most closely watched names in that hunt, each with a different strategy for capturing the market.
Broadcom has emerged as one of the strongest contenders because of its custom AI chip business, which is increasingly tied to large cloud customers building their own infrastructure. The company has said it sees a path to more than $100 billion in AI chip revenue by 2027, and recent reports suggest Broadcom's custom AI ASIC revenue is already far ahead of Marvell's. Broadcom also continues to benefit from major hyperscaler partnerships, reinforcing its position as a core supplier in AI data centers.
Marvell is another stock drawing attention, especially because of its role in networking and custom silicon for AI systems. While it trails Broadcom in scale, reports show Marvell has been winning design activity in areas such as high-speed interconnects and AI infrastructure, where efficiency and flexibility matter. That makes Marvell a more speculative but potentially higher-upside play if demand for AI networking continues to accelerate.
AMD is trying to challenge Nvidia more directly on compute. The company has spent the past year pitching its Instinct accelerator lineup as a lower-cost alternative for buyers that do not want to rely solely on Nvidia's ecosystem, and its AI strategy has gained fresh attention after major cloud and enterprise partnerships. But AMD still faces a steep climb because Nvidia's lead in software, performance and market share remains enormous.
Intel, meanwhile, is playing a different game. Rather than trying to match Nvidia in dominance right away, Intel is positioning itself around data center architecture, standards and chip ecosystem support, with a focus on CPUs, networking and CXL-related technologies that help AI systems scale. That gives Intel relevance in the AI buildout, but it is not yet viewed as the pure-play growth story investors usually associate with the phrase "next Nvidia".
The problem for all four stocks is that Nvidia's moat is still exceptionally wide. One recent market analysis said Nvidia holds roughly 85% to 92% of the AI chip market, which shows just how much ground competitors still have to cover. Even so, the size of the AI infrastructure market means there is room for multiple winners, especially among companies supplying custom chips, networking, memory and interconnects.
For investors, the best answer may depend on what kind of AI exposure they want. Broadcom looks like the strongest blue-chip infrastructure bet, Marvell offers a more aggressive networking and custom-silicon upside case, AMD is the direct challenger to Nvidia's compute lead, and Intel is the turnaround play with strategic AI relevance but less momentum.
The search for the next Nvidia is really a search for where AI spending will go next. Instead of betting only on the dominant GPU maker, investors are now weighing the chip suppliers, network builders and custom-silicon designers that could benefit as AI moves deeper into cloud data centers.
