From Intel Inside era to fighting for a comeback in the AI age

Once synonymous with the “Intel Inside” era that powered the global PC boom, Intel is now fighting to reclaim relevance in an industry it once dominated. As design-led players like Nvidia surge ahead in the AI era and manufacturing leaders such as TSMC set the pace on advanced nodes, Intel finds itself navigating one of the most critical phases in its history.

Recent developments suggest a shift in momentum. Strategic backing from the US government, a reported $5 billion investment in Intel shares by Nvidia and emerging partnerships tied to Elon Musk’s ambitious Terafab project have sparked cautious optimism. Add to that Intel’s role in supplying CPUs for Google’s AI infrastructure and the company appears to be stitching together a new playbook.

But for a company still rebuilding trust after years of execution missteps, it is far too early to call this a comeback.
 
Twin challenges, rising pressure

Intel’s struggle is playing out across two critical battlegrounds.

In AI chips, it has steadily lost ground to Nvidia, whose GPUs have become the backbone of the generative AI boom. At the same time, Intel’s foundry business, central to its IDM 2.0 (integrated device manufacturing) strategy, has been bleeding billions while struggling with delays, customer traction and credibility against established players like TSMC and Samsung Electronics.

Must read: Why Japan's semiconductor consolidation is a wake-up call for India's chip ambitions

Intel Foundry alone posted losses of more than $10.3 billion in 2025, underscoring the scale of the challenge. Semiconductor manufacturing is capital-intensive, time-consuming, and unforgiving on execution, areas where Intel has faltered in recent years.

A lifeline for Intel Foundry?

Much of the focus has turned to the Terafab project, but key details of the partnership remain unclear. What is evident, however, is that Intel’s role could extend beyond traditional manufacturing.

“Intel’s engagement with Terafab represents a strategic shift beyond traditional foundry services toward full-stack integration. The company is likely combining leading-edge manufacturing (18A/sub-2nm), advanced packaging, and design co-optimisation, positioning itself as a systems-level partner rather than a neutral capacity provider,” said Manish Rawat, semiconductor analyst at TechInsights.

Even so, the deal does not yet establish Intel Foundry’s credibility in the market.

Danish Faruqui, CEO at Fab Economics, pointed to supply constraints as a key driver behind the partnership. “Existing suppliers like TSMC and Samsung are running at capacity. Therefore for Musk’s vertical integration and leading edge supply security goals, Intel is the only partner with the scale and US location required for his ‘epic chip building exercise’,” he said.

Must read: West Asia crisis: A war far from fabs is rattling the chip supply chain

Terafab could emerge as an anchor customer and provide high-volume demand, but profitability remains distant. “Given the 3-5 years of fab construction timeline, this could only translate to compensating operating margins if US CAPEX subsidies support the geoeconomics of semiconductor value chain and Intel delivers BIC yields,” Faruqui added.

Finding relevance beyond GPUs

Even as its foundry ambitions remain under scrutiny, Intel is recalibrating its role in another critical layer of the AI stack.

Rather than competing head-on in AI accelerators, the company is leaning into its strengths in CPUs, positioning itself as a key enabler of AI infrastructure through its Xeon 6 processors.

In GPU-heavy AI systems, CPUs remain indispensable, handling task scheduling, managing data movement, running operating systems, and ensuring overall system orchestration. While GPUs accelerate compute-intensive workloads, CPUs keep the system functioning efficiently.

Must read: How Indian conglomerates are shifting India’s semiconductor playbook

Intel’s partnership with Google to supply CPUs for AI workloads reflects this shift. More notably, Nvidia has also opted for Xeon 6 CPUs in its DGX Rubin NVL8 systems, reinforcing Intel’s continued relevance in the broader compute stack.

“Concerns regarding the timely availability of GPUs and higher cost are driving companies towards optimising their spend on AI infrastructure and looking at a more heterogeneous computing infrastructure. It can optimise the AI spend, making it a win-win for all,” said Devroop Dhar, CEO and co-founder of Primus Partners.

He added, “This deal could be a game-changer for Intel and give it a much-needed boost in the AI infrastructure space, where it has lagged behind players such as Nvidia and AMD. Locking in a large hyperscaler such as Google would bring in a great deal of certainty into the operations of Intel.”

For companies like Google and Nvidia, the partnership also offers greater reliability in CPU supply amid global uncertainty, while helping optimise compute costs.

Execution will decide the outcome

Intel’s evolving role points to a broader strategic recalibration. The company may finally have the right partnerships and positioning, but the semiconductor industry has little patience for promises.

In a market defined by execution, not intent, Intel must deliver on timelines, yields and customer commitments. Failure to do so could mean missing what may well be its last major opportunity to reclaim leadership.