In the fourth quarter, AMD’s data center segment delivered a 34% quarter-over-quarter revenue jump to $4.3 billion, driven primarily by rapid enterprise adoption of its MI300 series accelerators. Operating income in the division surged 793% year-over-year as gross margins expanded on higher utilization of GPU capacity. The MI300 line now powers orders from three of the five largest global cloud providers, and backlog for MI400 pre-orders stands at more than $2.1 billion ahead of its 2026 launch window.

At CES 2026, CEO Lisa Su introduced the Helios rack-scale platform built around the new Instinct MI455X accelerator. Each Helios rack delivers up to 3 exaflops of AI performance in a single chassis, positioning AMD to compete directly with incumbent offerings. The event also highlighted the Ryzen AI Halo client processor, underscoring AMD’s strategy to extend AI inference workloads from hyperscale data centers to edge and workstation environments.

Since late October, AMD’s share price has retraced nearly 20% despite robust fundamentals. In the third quarter, the company reported a 5.6% revenue surprise, posting 35.6% year-over-year top-line growth that outpaced the preceding quarter’s 22% gain. Analysts have revised forward revenue estimates upward for four consecutive months, reflecting confidence in AMD’s long-term trajectory even as short-term sentiment cooled.

With MI400 series accelerators slated for general availability in mid-2026, AMD is targeting inference and hyperscale compute markets projecting a 42% compound annual growth rate through 2029. The company plans to double its data center R&D investment to support next-generation GPU architectures and is in talks with five leading telecommunication firms to deploy edge AI clusters. Consensus models forecast annualized AI data center hardware growth of 80%, underpinning an overall corporate revenue growth target of 35% per year over the next three years.