Advanced Micro Devices (AMD) has set a concrete deadline for its most ambitious growth target: the company's data-center business is expected to more than double by 2027. This commitment, reiterated by executives at the Goldman Sachs Communacopia + Technology Conference on September 11, is a far more demanding test than the broad artificial intelligence (AI) demand narrative that has driven the stock higher. Success hinges on translating an expanding product portfolio into rack-scale deployments and revenue growth on schedule.
During the conference, Matt Ramsay, AMD's corporate vice president of financial strategy and investor relations, reaffirmed the company's long-term goals of exceeding 60% compound annual growth for the data-center franchise and over 80% for its AI business. More immediately, he referenced a target of “much more than doubling” the data-center segment in 2027, according to a transcript of the session. This claim is grounded in a substantial base: AMD reported second-quarter data-center revenue of $6.7 billion, up 107% year-over-year, representing 58% of total company revenue of $11.54 billion. The segment includes both EPYC server processors and Instinct accelerators, so the target is not solely a GPU forecast.
The current quarter provides a reality check. Annualizing AMD's second-quarter revenue yields approximately $46.1 billion. Against the company's equity value of roughly $843 billion at Friday's close, the stock traded near 18 times that backward-looking run rate. While this is a rough calculation—revenue is growing rapidly, seasonality matters, and future margins are not accounted for—it underscores the high level of execution investors already expect.
AMD's server opportunity is expanding on management's spreadsheet as well as in reported sales. Ramsay said the company had increased its estimate for the 2030 server total addressable market (TAM) from $60 billion to $120 billion and then to $220 billion. He described a goal for AMD's server business to exceed half of that market, potentially building a $100 billion server operation. These figures are ambitions, not booked orders or formal segment guidance, but they reflect a broader thesis: agentic AI workloads require more host processors and general-purpose servers around accelerators, while AMD gains enterprise share with EPYC and broadens its AI offering from silicon to racks, networking, and software.
Dan McNamara, who leads AMD's compute and enterprise AI business, noted that enterprise server revenue grew more than 70% in the second quarter. However, management also acknowledged that the current AI business remains concentrated among its largest customers. For 2027, the milestones are more concrete than the TAM: investors can watch whether Helios rack deployments scale, whether the Venice server launch reaches cloud and enterprise buyers on time, and whether ROCm software reduces the effort required to deploy AMD systems. A rising segment revenue line without improving economics would be a weaker outcome. Ramsay said AMD expects gross-margin dollars to grow significantly faster than expenses as the data-center business expands.
The target bundles several hard jobs into one year. AMD must secure advanced manufacturing and memory, assemble complete systems at scale, keep its product cadence, deepen its software stack, and compete with Nvidia, Intel, and customers' in-house silicon. The company's latest quarterly filing lists export controls, manufacturing capacity, component availability, customer concentration, and competitive product timing among the risks that can change results.
There is also a distinction between a conference target and quarterly guidance. AMD's current formal outlook covers only the third quarter: approximately $13 billion of revenue, plus or minus $300 million, and about 56% non-GAAP gross margin. Until the company provides a 2027 segment bridge, the doubling claim cannot be translated cleanly into a revenue number or earnings estimate.
The next decisive evidence will be the conversion of announced rack-scale partnerships and new server platforms into reported data-center revenue while gross-margin dollars outpace operating costs. AMD has now provided the deadline. The burden shifts to quarterly results.
