Intel unveiled Xeon 6+ CPUs aimed at high-density, scale-out and inference-heavy workloads, signaling a renewed push into AI data-center infrastructure even as GPUs remain central to acceleration.
According to CPU Benchmark, Intel holds about 60% of the global CPU market and Advanced Micro Devices about 38.5%, while most estimates place Nvidia at a minimum 80% share of the AI accelerator market; AMD reported revenue growth of 34% in 2025 and 38% in the first quarter of 2026 as Intel's top line rose 7% in Q1, and AMD projects the data-center CPU market will grow at an annualized 35% through 2030 to $120 billion.
Intel's recovery is set against a decade of manufacturing and foundry setbacks-10‑nanometer scale-up problems in 2015, a stalled 7‑nanometer push in 2020, and a scaled-back foundry expansion-while AMD has leaned on Taiwan Semiconductor Manufacturing and invested in both capacity and product performance, and as then-new Intel CEO Lip‑Bu Tan said in 2025, "Over the past several years, the company invested too much, too soon."
AMD also plans product rollouts that test this thesis: its MI450 GPUs, which the company says offer up to 432 gigabytes of high-bandwidth memory and 19.6 terabytes per second of memory bandwidth, are slated to debut later this year.
