Morgan Stanley, one of the world's largest investment banks, has issued a bullish outlook for Cerebras Systems, the AI chipmaker that has carved out a niche in the fast-growing market for AI inference. The bank projects that Cerebras' core revenue will more than triple by 2027, driven by a surge in demand for the computing power that runs AI models after they've been trained.
AI inference is the process where a trained AI model applies what it has learned to answer real-world questions—like generating a response in a chatbot or recognizing an object in an image. It's the stage that happens after the expensive and time-consuming training phase, and it's where most businesses actually interact with AI. Right now, demand for inference computing is outstripping the available supply, creating a bottleneck that companies like Cerebras are trying to fill.
What Morgan Stanley is saying
Morgan Stanley's analysts believe Cerebras can close that gap by winning more large cloud customers and expanding the physical infrastructure that runs these workloads. The bank now expects Cerebras' 2026 core revenue to reach $886 million, up from earlier estimates, and sees the company's revenue trajectory accelerating sharply through 2027.
The key to that growth, according to the bank, is Cerebras' ability to scale its data-center capacity. The company is planning to bring 600 megawatts of new capacity online—a massive undertaking that involves building or leasing facilities, installing power and cooling systems, and deploying thousands of specialized chips. That's the kind of project that can make or break a young hardware company, and Morgan Stanley flags it as the biggest execution test.
Cerebras is also expanding its product lineup. The company is developing new inference systems built in partnership with Advanced Micro Devices (AMD) and Amazon Web Services (AWS). These collaborations are designed to give Cerebras access to a broader range of customers and to integrate its technology into the cloud platforms that many businesses already use.
Why AI inference matters
To understand why this is a big deal, it helps to know the difference between AI training and inference. Training is the phase where an AI model is fed massive amounts of data to learn patterns—think of it as the model's schooling. Inference is when the model is put to work, answering questions or making predictions in real time. Training is a one-time (or periodic) cost, but inference happens every time someone uses an AI tool, which means it's a recurring and rapidly growing source of demand.
That's why companies like Cerebras are betting big on inference. While Nvidia dominates the overall AI chip market, Cerebras has focused on a specific slice: building chips that are particularly fast at inference tasks. The company's wafer-scale engine, which is a single chip the size of a dinner plate, is designed to handle large AI models quickly and efficiently.
The broader AI infrastructure boom has been a major theme in markets over the past year. Tech giants and cloud providers are spending billions on data centers and chips, and that spending has lifted the stocks of many semiconductor companies. But it's also created a crowded field, and investors are increasingly focused on which companies can actually deliver on their growth promises.
What it means for investors
For everyday investors, Morgan Stanley's report is a signal that Cerebras is seen as a credible player in the AI infrastructure race, but it also highlights the risks. The company's growth depends on executing a massive buildout of data-center capacity—a process that can be delayed by supply chain issues, permitting problems, or technical hiccups.
If Cerebras succeeds, the payoff could be substantial. The bank's projection of more than tripling core revenue by 2027 implies a company that is growing at a breakneck pace, which could justify a high valuation. But if the capacity expansion falls behind schedule, the company could miss those targets, and its stock could suffer.
It's also worth noting that Cerebras is not yet profitable on a core basis, and its revenue is still relatively small compared to giants like Nvidia or AMD. The company's future depends on its ability to convert its technological advantages into sustained commercial success.
For context, other companies are also riding the AI wave. Lenovo's AI revenue surged 43% recently, though a derivatives loss dragged its bottom line. And Cisco's 2027 revenue forecast fell short of AI-driven hopes, showing that even established tech firms face high expectations. Cerebras, by contrast, is a pure-play AI chipmaker, which makes it more volatile but also more directly tied to the AI boom.
Morgan Stanley's report also comes on the heels of Cerebras raising its own 2026 forecasts, as AI inference demand accelerates. The company has been adding customers and expanding its product line, and the bank's analysis suggests that momentum will continue.
The bottom line
Morgan Stanley's bullish view on Cerebras is a bet on the continued growth of AI inference. The company has a unique technology and a clear strategy, but the real test will be whether it can build out 600 megawatts of capacity without major delays. For investors, that's the number to watch.
As always, it's important to remember that analyst forecasts are just one opinion, and they can be wrong. The AI market is evolving rapidly, and competition is fierce. But for those looking to understand the AI infrastructure landscape, Cerebras is a name that's likely to keep appearing in the headlines.
