Intel is making a new push into GPUs, this time with a focus on data center workloads, as the chipmaker looks to reestablish itself in a market increasingly shaped by AI-driven demand and dominated by Nvidia.

CEO Lip-Bu Tan said that after hiring a senior GPU architect, the company is working directly with customers to define requirements, signaling a more demand-driven approach as enterprises and cloud providers weigh their options for accelerated computing, according to a Reuters report.

Intel’s push comes as demand for AI accelerators reshapes data center spending, leaving enterprises and cloud providers with fewer GPU options and longer procurement timelines.

This is not Intel’s first foray into discrete graphics. The difference now is that it’s tying its GPU ambitions more closely to its data center roadmap and broader manufacturing strategy, pairing closer customer engagement with advanced process technology to gain traction.

Intel’s enterprise advantage

Intel’s tight integration of CPUs, GPUs, networking, and memory coherency gives it an edge in enterprise inference, hybrid cloud, and regulated or on-prem environments, where cost control and operational simplicity matter more than peak performance, said Manish Rawat, a semiconductor analyst at TechInsights.

In these segments, Intel has an opportunity to meaningfully limit Nvidia’s expansion and reduce customer dependence at the infrastructure level.

Supply chain reliability is another underappreciated advantage. Hyperscalers want a credible second source, but only if Intel can offer stable, predictable roadmaps across multiple product generations.

However, the company runs into a major constraint at the software layer.

“The decisive bottleneck is software,” Rawat said. “CUDA functions as an industry operating standard, embedded across models, pipelines, and DevOps. Intel’s challenge is to prove that migration costs are low, and that ongoing optimization does not become a hidden engineering tax.”

For enterprise buyers, that software gap translates directly into switching risk.

Tighter integration of Intel CPUs, GPUs, and networking could improve system-level efficiency for enterprises and cloud providers, but the dominance of the CUDA ecosystem remains the primary barrier to switching, said Charlie Dai, VP and principal analyst at Forrester.

“Even with strong hardware integration, buyers will hesitate without seamless compatibility with mainstream ML/DL frameworks and tooling,” Dai added.

Lian Jye Su, chief analyst at Omdia, said Intel will need to focus on delivering performance and software that are accepted and certified by the developer community.

While CUDA dominates with extensive libraries, tools, and developer mindshare, developers may be willing to adopt Intel GPUs if the company “can create a GPU that can provide tools and SDKs that are developer-friendly and address cutting-edge AI applications,” Su added.

From an enterprise buying perspective, this means Intel’s challenge is less about hardware ambition and more about overcoming deeply entrenched platform lock-in.

“Performance and pricing advantages alone will fall short without seamless developer tools and broad compatibility,” said Prabhu Ram, VP of the industry research group at Cybermedia Research. “Even with tight GPU-CPU-networking integration offering efficiency gains, CUDA’s entrenched lock-in remains the major barrier for enterprises that seek to reduce reliance on Nvidia.”

Rising China challenge

The rise of Chinese alternatives adds urgency to Intel’s effort to reestablish itself as a credible second source for Western enterprises.

In the Reuters interview, Tan said he was surprised to see Huawei hiring top-tier chip designers despite US restrictions on access to advanced tools, warning that China could leapfrog established players if Western companies are not careful.

“Huawei’s significance isn’t about near-term benchmark parity, it’s about trajectory,” Rawat said. “Progress on EDA independence may be slow, but directionally it’s real. High talent density is compensating for tool gaps, while parallel “good-enough” design flows steadily dilute the effectiveness of US choke points.”

According to analysts, Huawei does not need to outperform Nvidia globally to pose a strategic challenge. Locking in China’s domestic data center demand, reducing dependence on Western supply chains, and building closed-loop learning and optimization cycles at home could be enough to reshape competitive dynamics over time.