Advanced Micro Devices eyes $94B revenue by 2027, nearing $100B threshold
AMD's AI-driven data center business has accelerated its own timeline by three years, with 2027 revenue now projected to approach $94 billion.
Three years ago, AMD’s $100 billion revenue ambition was the kind of goal companies pin to a distant horizon to sound impressive at investor days. Now it looks like a realistic near-term target, and the semiconductor company is moving faster than even its own analysts expected.
During AMD’s Q3 2026 earnings call in early November, CEO Lisa Su outlined a significant upward revision to the company’s 2027 revenue outlook, with projections now pointing to approximately $94 billion for that year. That figure implies over 80% growth year-over-year from a 2026 baseline forecast of approximately $52.2 billion.
How the data center became AMD’s whole story
The engine behind this revised outlook is not CPUs or gaming chips. It is the data center, which is expected to account for more than 80% of AMD’s total revenue.
Server CPU revenue is projected to roughly double as AMD expands customer deployments of its AI-focused processor lineup. But even that impressive trajectory is expected to be outpaced by GPU sales, with data-center AI revenue growing at a significantly faster rate than its CPU counterpart.
The adoption of AMD’s Instinct MI-series GPUs and its Helios rack-scale systems is driving the acceleration. Major AI developers including OpenAI, Meta, and Anthropic have reportedly secured multi-gigawatt deployments of Helios systems, with initial shipments beginning in late 2026.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
A timeline that moved three years forward
Context matters here. At AMD’s November 2025 Analyst Day, the company set a target of $100 billion in data-center revenue by 2030. The current projections suggest that milestone could arrive at the total company level by 2027, three years ahead of schedule.
The company is not without constraints, however. AMD has cited supply limitations, particularly around high-bandwidth memory, as a principal near-term ceiling on growth. High-bandwidth memory is a specialized component that sits alongside AI chips in data-center servers, and production capacity has not kept pace with demand across the industry.
What it means for the competitive landscape
For investors, the revised outlook carries meaningful implications. If the company executes on approximately $94 billion in 2027 revenue, the upward revision to earnings estimates could be substantial, particularly given how much operating leverage a data-center-heavy mix provides compared to consumer segments.
The broader question is whether the supply constraints around high-bandwidth memory resolve quickly enough to let AMD capture the demand it has already lined up. Customer deployments at the scale of OpenAI, Meta, and Anthropic require sustained hardware availability over multiple quarters, not a single shipment. Any elongation of delivery timelines risks pushing a portion of that revenue into 2028.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.

