Intel Cuts Legacy Jobs, Pivots to AI Chips

Intel's Painful Pivot to AI

Intel Corporation has announced a sweeping restructuring that will eliminate approximately 4,500 positions across its legacy hardware divisions while simultaneously creating 2,000 new roles focused on AI chip development. The move, affecting roughly 4% of Intel's global workforce, represents the most significant organizational shift since Pat Gelsinger's departure and the appointment of new CEO Michelle Johnston Holthaus.

The layoffs primarily target engineers and managers in Intel's client computing division (traditional PC processors) and its network and edge computing group. Meanwhile, the company is aggressively hiring for its Accelerated Computing Systems and Graphics (AXG) group and its foundry services division.

The Decline of Legacy Hardware

Intel's decision reflects a harsh market reality: the traditional CPU market that built Intel's empire is shrinking in both volume and strategic importance. Several factors have accelerated this decline:

• PC market contraction: Global PC shipments have declined for the fifth consecutive quarter, down 8% year-over-year
• ARM competition: Apple's M-series and Qualcomm's Snapdragon X chips have captured significant market share in laptops
• Data center shift: Cloud providers are increasingly deploying custom silicon and AI accelerators instead of general-purpose CPUs
• Margin pressure: Intel's average selling prices for client CPUs have dropped 12% as competition intensifies

"The era of the general-purpose CPU as the center of computing is ending. Intel either transforms into an AI-first company or becomes irrelevant. There is no middle path." — Former Intel executive speaking to TechCrunch

Doubling Down on AI Accelerators

Intel is channeling $8 billion in R&D spending toward its AI chip portfolio over the next two years. The centerpiece of this strategy is the Gaudi line of AI accelerators, which Intel acquired through its 2019 purchase of Habana Labs.

The upcoming Gaudi 4, expected to tape out in late 2026, represents a ground-up redesign that Intel hopes will close the performance gap with NVIDIA's Blackwell architecture. Key improvements include:

• Chiplet design: Modular architecture allowing flexible configurations for training and inference
• HBM4 integration: Next-generation high-bandwidth memory for larger model support
• Software ecosystem: Significant investment in OneAPI and open-source tooling to lower the barrier to migration from CUDA
• Competitive pricing: Intel plans to undercut NVIDIA by 30-40% on price-performance for inference workloads

Impact on Affected Workers

Intel has stated that affected employees will receive severance packages including 60 days of paid notice, continued health benefits for six months, and career transition support. The company is also offering retraining programs for employees who want to transition into AI-related roles within Intel.

However, the reality is that many of the eliminated positions require deep expertise in x86 architecture and traditional CPU design — skills that don't directly translate to AI accelerator development. Industry observers note that this creates a difficult transition for mid-career engineers who have spent decades specializing in legacy architectures.

For engineers navigating this transition, whether seeking new roles at Intel's AI division or elsewhere, strong interview preparation is essential. InterviewAlly helps tech professionals prepare for the system design and technical interviews that AI hardware companies prioritize in their hiring processes.

The Foundry Play

A crucial but often overlooked aspect of Intel's pivot is its foundry services business. Intel is positioning itself as a domestic manufacturing alternative to TSMC for AI chip production, leveraging $20 billion in CHIPS Act funding to build new fabs in Ohio, Arizona, and Oregon.

"We don't just want to design AI chips — we want to manufacture them. Having advanced fabrication on American soil is a strategic asset that becomes more valuable every year as geopolitical tensions rise." — Intel's new CEO in a recent earnings call

This dual strategy — designing its own AI accelerators while manufacturing chips for others — could give Intel multiple paths to relevance in the AI era. Several AI chip startups have already signed agreements to use Intel's 18A process node, providing both revenue and process validation.

What This Means for the Industry

Intel's restructuring is part of a broader pattern across the semiconductor industry, where companies are rapidly shifting resources from legacy computing to AI. AMD has similarly reduced investment in traditional CPU development, while NVIDIA continues to grow its workforce almost exclusively in AI-related roles.

For tech workers, the message is clear: AI skills are no longer optional, even in hardware engineering. The semiconductor companies that survive the next decade will be those that successfully pivot to AI, and they'll need engineers who can make that transition with them. The estimated 4,500 displaced Intel workers will join a growing pool of experienced engineers looking for roles in an industry that increasingly demands AI expertise.