Cisco Chief Product Officer Jeetu Patel presents at the recent Cisco Partner Summit.
One tech giant after another has waded into the AI waters in the past couple of years, but until recently Cisco’s AI strategy was a bit of a mystery to some. However, after attending the Cisco Partner Summit in late 2024, reviewing the company’s fiscal first quarter 2025 earnings, and having numerous conversations with Cisco executives, I can clearly see how the company has bided its time. to clarify a product and market strategy that aims to capitalize on its strengths in networking, security, data management and even computing to give Cisco a place prominent in enterprise hyperscale and AI infrastructure.
Mind you, it’s early. Last quarter, the company earned $300 million from specific AI products and says it’s on pace to reach $1 billion in AI revenue in fiscal 2025. That’s a strong start, though it should be understood in the context of a The company that it is has been over $50 billion in annual revenue for years now, so it also has plenty of room to grow. However, Jeetu Patel, who was named the company’s chief product officer last summer, has articulated a strategy aimed at giving Cisco a “platform advantage” where the company’s various offerings play together inter- functional for any customer – hyperscale or enterprise – looking to get the most from AI in the data center. This looks to me like Cisco is sticking to what it does best: connected platforms.
Building an AI strategy on Cisco’s existing strengths
The example of a platform advantage Patel cited during his Cisco Partner Summit keynote was his preference for Apple products: He’s had an iPhone for years, which plays nicely with his MacBook and every other personal electronic device like an iPad or Apple Watch. (I do the same with Samsung Galaxy products.) The point is that investing fully in one or two of those products makes it more likely that you’ll choose other products from the same manufacturer.
Of course, the math and purchasing decisions are quite different in enterprise B2B technology, but Patel makes a strong point about how well Cisco is positioned for this approach with enterprise AI. Cisco’s stable status in networking seems unbreakable; has been a major player in cybersecurity for decades and is showing renewed vigor in that area; its reach in data management (including monitoring and security) has only grown through its acquisitions in recent years – most notably the acquisition of Splunk; and its ambitions in high-end data center computing are at least plausible. Importantly, many enterprises are as connected to Cisco products in the data center as Patel is to his Apple products or I am to my Samsung devices.
That said, I was a little surprised to hear Patel say during his keynote that “We’re going to double down on the computer business” and “We’re unapologetically in the computer business.” It absolutely makes sense from a platform perspective to provide a full stack, both for the technical ease of interoperability and the business advantages of vendor loyalty. Plus, Cisco has traditionally been good at end-to-end experiences. But I’ve never heard this kind of computer talk from Cisco before.
There are limits to where Cisco can or should try to go with computing. In particular, the margins in computing are much lower than in any of Cisco’s other businesses, so in my opinion the company would not thrive if it tried to compete in data center computing at scale with Dell, Lenovo and HPE. I think it’s going to be a matter of choosing the right niches to compete on and adding computing together with other products where it makes sense to round out the offering to a specific customer. The company’s blades and its “better together” philosophy work well for those customers who appreciate that approach — and Cisco needs to find more customers who do. Enterprise AI is complex, and that’s why this focus is smart.
Enabling loads of AI at scale
With all of this as context, two big infrastructure products that Cisco has announced make a lot of sense for the AI data market. The Nexus 9000 switch comes from the deepest part of Cisco’s core competency; is a highly scalable and efficient 800 gigabit switch already used by hyperscalers. And Cisco is using its Unified Computing System approach—which combines compute, networking, and storage into a single system—to provide a complementary server with eight Nvidia GPUs for AI training. I consider UCS to be one of Cisco’s flagship “light button” offerings, and the combination of these two new products supports Patel’s claim that “We are now in the AI infrastructure business.”
The most profitable area in AI over the past couple of years has been training, but Patel and Cisco are also clearly focused on increasing the inference of artificial intelligence and the deployment of AI applications in the enterprise, which should begin to grow in earnest in 2025 , likely in the second half of the year. Admittedly, it was great to learn from the recent earnings call that Cisco has scored at least one major design win for its AI hyperscaler and sees continued momentum in recent networking for LLM hyperscaler training groups. But Patel notes that about $200 billion has been spent industry-wide on AI training so far to bring in about $5 billion to $10 billion in revenue. So we must think that a greater reward is coming. As enterprises—which are rarely the first movers in new areas of technology—move from experimentation to mass deployment of truly needle-moving AI applications, we can expect much more revenue to come from inference and AI. of the enterprise in general. Cisco is positioning itself to capture a significant portion of IT spending that will drive those revenues.
In line with this, Cisco is infusing AI into all its products. As Patel told the audience of Cisco partners during his keynote, “You should expect that every product we build . . . there is artificial intelligence built into the fabric and the way we think about building the product. It’s not an afterthought.”
The next “easy button” from Partner Summit is its AI PODs for inference, which are plug-and-play infrastructure clusters—in this case using Nvidia software—configured for specific industries and use cases. As with UCS, these products combine compute, networking and storage, plus they add cloud management functions. They are built to be scalable and very quick to replicate, which should increase their appeal to enterprises.
Another thing struck me when my business partner Daniel Newman and I interviewed Patel for the Six Five on the Road segment: He clearly understands the fundamental opportunity in big, untapped enterprise data that’s not yet being used to train AI models . It seems like forever I’ve been reminding people about the vast majority—perhaps 80%—of enterprise data that is NO in the cloud and NO available for LLMs based on publicly available data. This is the data that smart early adopters among enterprises are using to adapt and fine-tune their internal models. This is also the data that smart vendors from SAP to ServiceNow to AWS to Microsoft to IBM are helping their customers use for AI. Cisco now seems intent on enabling what Patel calls “boring” functions in edge networks, model security and so on that will enable many enterprises to make the most of this data with AI.
Selling picks and shovels to AI Gold miners
Given Cisco’s size and the complexity of its portfolio, there’s a lot more I could say about its efforts in AI training infrastructure, AI connectivity, and AI inference. At some point I may also do a deeper dive into its Silicon One initiative, which helps Cisco incorporate its own silicon into its devices. This is likely to become more important over time in the AI realm, because neither OEMs nor end customers are fully satisfied with a market in which only one or two semiconductor manufacturers (read: Nvidia, on the one hand of AMD) dominate the scene.
The most important thing right now is to understand how Cisco is positioning itself in the AI market, especially for enterprise customers. Patel likes to point out that the people who probably got rich during the California Gold Rush weren’t prospectors and miners, but suppliers who sold picks and shovels. We already saw this with hyperscalers, and I believe it will be true in the enterprise data center as well. The implication of his analogy is clear: Whatever hyperscalers or, increasingly, enterprises achieve or fail to achieve with generative AI, Cisco can benefit from these customers’ need for fast, scalable, and secure network architecture that supports AI training, association and inference. . For a larger company like Cisco, the challenge is always execution, but there’s no doubt in my mind that the opportunity is there.
While I’m not yet sure how material Cisco’s computing efforts will be for its AI strategy—I need to see some real-world results before I’m a believer—I have to give the company credit for a smart strategy that builds on her existing portfolio and relationships, and smart moves from past years. In particular, I would point out the wisdom of its decisions years ago to integrate Nvidia AI into Cisco’s collaboration devices. (I have one of these devices sitting next to me on my desk as I write this.) The quality, security, and user experience that Cisco provides is consistent, and Patel is right to say that he considers quality to be “zero priority.” , which means he aims to be such that people don’t even talk about him anymore.
With the acquisition of Splunk and the main inhibitor of enterprise AI being data management, I’d like to see the company double down on a full-scale data management platform. By “full-scale,” I mean one that can compete with Cloudera, Databricks, and Snowflake.
As an ex-product guy, I’m biased, but I really like Patel’s product philosophy. He wants to build great products that customers love, and he wants Cisco to focus on the “10x” market opportunities where Cisco can outrun the competition. Following the lead of his CEO, Chuck Robbins, he also knows that “Tempo Matters.” In other words, now that Cisco has taken time to fine-tune its AI product mix and approach to market, it intends to move QUICK. I think 2025 will provide Cisco with plenty of opportunities to prove the wisdom of this approach.
