Arm (NASDAQ: ARM) Neoverse & AWS Graviton4 vs. AMD EPYC Turin: Will Custom Cloud CPUs Break the x86 Server Duopoly?
Built on TSMC (NYSE: TSM) 3nm nodes and Arm Holdings (NASDAQ: ARM) Neoverse cores, custom hyperscaler processors like AWS Graviton4, Google Axion, and Microsoft Cobalt 100 now capture 10% to 15%+ of global server shipments, while AMD's 192-core EPYC Turin reaches 40%–46% x86 revenue share.
Per SEC Form 20-F and Form 10-Q disclosures and Mercury Research trackers, agentic AI orchestration and head-node pre-processing have ignited a host-CPU shakeup across JEDEC DDR5 and PCIe 6.0 racks. While Advanced Micro Devices (NASDAQ: AMD) squeezes Intel Corporation's (NASDAQ: INTC) Xeon 6 Granite Rapids to 54%–60% unit share, Amazon (NASDAQ: AMZN) runs over 50% of new AWS compute additions on Graviton4.
For investors, Arm's shift toward pre-integrated Armv9 Compute Subsystems (CSS) roughly doubles per-chip royalty rates while offering cloud giants 30%–40% price-performance gains over merchant x86 silicon. Will custom Arm Neoverse chips or AMD's 192-core EPYC Turin capture more AI data center CPU budget?
Which architecture will win the AI data center host CPU market by 2027?
- Arm Neoverse and AWS Graviton4 will take 25%+ cloud CPU share0 %
- AMD 192-core EPYC Turin will dominate enterprise x86 servers0 %
- Depends on Armv9 CSS royalty costs and Intel Xeon 6 yields0 %
Comments
Be respectful. Comments reported by several people are hidden until a moderator reviews them.
You do not need an account to read, only to comment. Sign in with a code sent to your email or with your Google account. We only show the name you choose, never your email.