TLDR
- Jensen Huang announced that Nvidia is prepared to challenge Intel and AMD directly in the CPU space
- The AI industry is transitioning from GPU-intensive model training to CPU-optimized agent deployment
- Nvidia introduced its Grace and Vera processor lineup in 2023, optimized for data center throughput
- Meta Platforms secured a standalone CPU agreement with Nvidia, marking the first time these chips will ship without accompanying GPUs
- Competition intensifies as AMD simultaneously announced its own separate CPU partnership with Meta
For years, Nvidia’s revenue stream has been driven primarily by graphics processing units. Now, CEO Jensen Huang is signaling a strategic expansion into central processing units, a domain where Intel and AMD have reigned supreme.
During Wednesday’s fourth-quarter earnings discussion with analysts, Huang articulated Nvidia’s readiness not merely to participate in the CPU resurgence, but to dominate it.
“We love CPUs as well as GPUs,” Huang declared during the analyst call.
For many years, central processing units served as the backbone of computational tasks. Graphics processors subsequently captured significant market share as artificial intelligence model training surged, demanding the parallel processing capabilities that GPUs excel at delivering.
However, the computational landscape is evolving once more. AI enterprises are transitioning from model development to model deployment — implementing what the industry calls “agents” that perform coding, document analysis, and report generation.
These agent-based workloads are better suited to CPU architecture, according to Creative Strategies analyst Ben Bajarin. He noted that agentic computing “is happening more and more, and sometimes primarily, on the CPU.”
The CPU vs GPU Shift in AI
Nvidia’s present top-tier AI server configuration, the NVL72, contains 36 central processors alongside 72 graphics processors. Bajarin projects this ratio could evolve toward parity — or that certain workloads might eliminate GPU requirements altogether.
Nvidia unveiled its Grace and Vera central processing units for data center applications in 2023. Huang explained these processors employ a fundamentally different architecture compared to Intel and AMD offerings, prioritizing data throughput over general-purpose versatility.
“It is designed to be focused on very high data processing capabilities,” Huang explained.
Nvidia recently finalized an agreement with Meta Platforms for substantial CPU volumes delivered independently — representing the first time these processors will ship without their traditional GPU counterparts. This departure signals a fundamental change in Nvidia’s go-to-market approach.
Meta isn’t abandoning its current CPU vendors. Instead, the social media giant is diversifying its supplier base by incorporating Nvidia. Shortly following the Nvidia announcement, AMD disclosed its own CPU supply agreement with Meta.
Intel’s Dominance No Longer Guaranteed
HotTech Vision and Analysis analyst Dave Altavilla observed that Nvidia aims to demonstrate the CPU “is no longer the assumed default foundation of modern compute infrastructure.”
At CES in January, Huang predicted that Nvidia CPU adoption in data centers would “explode,” expressing confidence that Nvidia could emerge as “one of the largest CPU makers in the world.”
Nvidia intends to unveil additional details regarding its CPU development roadmap at its upcoming annual developer conference in Silicon Valley next month.
