Nvidia is laying the groundwork to expand meaningfully in the data-center CPU market, an area long dominated by Intel and Advanced Micro Devices. The company’s shift is tied to accelerating demand for AI compute, as more workloads move from GPUs alone to CPU-assisted processing across training and inference systems.
In its fiscal Q2 2027 reporting, Nvidia said its Grace CPU family generated $5 billion in sales over the trailing 12-month period, and it forecast that data center CPU revenue would more than double from the current fiscal year to the next, supported by the rollout of its next-generation processor lineup. The move comes alongside production scale-up for the company’s Vera CPUs and pairing with its high-end Vera Rubin GPUs.
Key takeaways
- Growth outlook: Nvidia projects data center CPU revenue would more than double from this fiscal year to next.
- Catalyst: Demand for AI compute is increasingly pulling CPUs into the workload mix, while Nvidia scales production of its Vera CPUs and rolls them out with Vera Rubin GPUs.
- Competitive implication: Nvidia aims to win share from Intel and AMD in server-class CPUs by offering systems designed to deliver integrated performance.
- Strategic shift: The company is positioning itself as a broader compute provider, not just a GPU leader.
What drove Nvidia’s push into server CPUs
Nvidia’s historical advantage has centered on GPUs used for artificial intelligence workloads, particularly for the heavy compute requirements of training models. Over time, however, the company’s opportunity has expanded as more AI processing is being shifted to CPUs as part of end-to-end data center architectures.
According to Nvidia’s fiscal Q2 2027 report, the Grace family is already contributing to the server CPU opportunity. Nvidia also signaled that the next phase of growth depends on scaling its broader CPU platform, including the introduction of its next-generation processor family.
The company’s plan is not limited to selling CPUs in isolation. Nvidia has been moving toward a more complete solution approach, combining hardware with its broader ecosystem and leveraging its ability to optimize performance across CPU and GPU components inside a data-center deployment.
Role of Vera CPUs and integrated performance claims
Nvidia said it has entered full production of its Vera CPUs. The company also indicated that the strategy includes using its Vera CPUs alongside Vera Rubin GPUs, with an emphasis on selling the technology as a stand-alone product where customers want CPU capacity tailored to AI data center needs.
Nvidia further stated that Vera hardware is designed to deliver performance and efficiency that significantly exceed those of other data center CPUs in key areas. The company’s argument is that this advantage could translate into faster market share gains versus Intel and AMD, particularly as data center operators look to optimize both throughput and power efficiency for AI workloads.
Market reaction: what investors are likely watching
While the report describes Nvidia’s outlook rather than immediate market pricing, investors typically interpret such CPU expansion plans through a few lenses: whether Nvidia can translate GPU dominance into CPU acceptance, how quickly it can scale production, and whether its CPU performance claims hold up in deployed systems.
For Intel and AMD, the key question is whether Nvidia’s integrated approach reduces the competitive foothold traditionally held by x86 server platforms. For Nvidia, the implication is that it may broaden its revenue base beyond GPUs and potentially capture additional spending from data-center operators building AI clusters.
Another consideration for investors is that CPU growth can be more dependent on broad compatibility across server vendors and customer platforms than GPU shipments. Nvidia’s messaging points to an advantage from optimization for integrated performance, but the pace of adoption in the server CPU market is ultimately shaped by customer qualification cycles and system-level integration.
Bigger picture: AI compute is becoming more CPU-inclusive
Nvidia’s strategy reflects a broader shift in AI infrastructure. As AI systems scale, data centers increasingly rely on balanced compute stacks rather than a GPU-only approach. CPUs are often tasked with orchestration and additional compute alongside accelerators, and demand growth in both training and inference can increase the value of server CPU performance and efficiency.
By forecasting that data center CPU revenue will more than double from this fiscal year to the next, Nvidia is signaling confidence that the market will expand and that its product roadmap will capture a growing share of spending. The company’s ability to scale its Grace platform and deliver Vera CPUs at production levels are central to that view.
Looking ahead, investors will likely focus on Nvidia’s next reporting cycle for updates on CPU revenue trajectory, margin trends, and guidance around the pace of Vera rollout. Additional signals from data-center hardware deployments—along with broader AI capex trends and any changes in enterprise and cloud demand—could determine whether Nvidia’s CPU ambitions translate into sustained share gains versus Intel and AMD.
