Morgan Stanley has revised its outlook on major U.S. chipmakers, raising its price target for Broadcom Inc. (NASDAQ: AVGO) to $443 from $409 and for Nvidia Corp. (NASDAQ: NVDA) to $250 from $235, while reiterating an ‘Overweight’ rating on Nvidia.
The revisions come amid mounting evidence that artificial intelligence (AI) demand is pushing the semiconductor supply chain to its limits, based on findings from a recent Asia trip and further industry checks.
According to Morgan Stanley, AI strength is pervasive across the sector, prompting upward revisions not only for Broadcom but more significantly for Nvidia, as both companies are expected to see substantial growth in 2026. The analyst also notes robust trends in memory markets and opportunities from China localization prospects.
The brokerage firm maintains that Nvidia is likely to retain its dominant market share, with competition in the AI chip space considered ‘overstated.’ The primary challenge for customers in the next twelve months will be securing adequate supplies of Nvidia’s products, especially the Vera Rubin line.
While market participants continue to seek alternatives to Nvidia, the bank’s checks reveal that even Google’s TPU—widely touted as a strong competitor—cannot fully displace Nvidia, as the latter still meaningfully contributes to data center workloads, including those that leverage TPU. Nvidia’s data center revenues recently stood at $51 billion for the last quarter, substantially surpassing those derived from TPU deployments.
Morgan Stanley’s raised estimates for Nvidia come after extensive meetings with industry contacts in Asia and the United States. Although Nvidia’s new estimates do not yet reach the ambitious ‘$500 billion in five quarters’ revenue trajectory set by its CEO, the outlook remains robust, with conviction in the continued momentum of Nvidia’s AI platform.
Responding to questions about Google’s TPUs, Nvidia CEO Jensen Huang stated, “We’ve been competing against ASICs for a long time. Google has had them for a long time, and they do a great job. What Nvidia does is much more versatile. We are everywhere. We are in every cloud. The Nvidia opportunity is just much larger.”
Huang emphasized that while Google’s TPUs are strong as specialized ASICs, Nvidia’s GPUs—along with its CUDA software—enable a unified and comprehensive AI ecosystem that TPUs cannot replicate.
