Intel Shares Jump on Google and Nvidia Reports: The Details

Intel shares (INTC) have jumped 12.84% to $111.90 on Monday after reports emerged that Alphabet’s Google has placed an order for more than three million specialized AI chips and Nvidia is testing the chipmaker’s advanced manufacturing technology, marking a potential breakthrough for Intel’s struggling foundry business.

The rally extends Intel’s remarkable year-to-date performance to 182%, transforming the stock from a turnaround story into one of 2026’s standout semiconductor plays. It’s worth noting that INTC had fallen back 11% on Friday, with today’s pop more than recapturing the $110 level on the day.

Core Details of the Google Partnership

According to multiple reports, Google has committed to ordering more than three million Tensor Processing Units from Intel for delivery in 2028. These TPUs are Google’s proprietary AI accelerators, central to powering the company’s Gemini AI models and broader cloud infrastructure.

The decision follows months of testing by Google to validate Intel’s advanced manufacturing capabilities. Morgan Stanley analysts estimate Google’s total TPU production across 2027-2028 could reach six million units or more, suggesting Intel may capture roughly half of Google’s manufacturing needs for that period. This represents a significant shift from Google’s historical reliance on Taiwan Semiconductor Manufacturing Company for cutting-edge chip production.

The strategic implications extend beyond a single customer. Industry sources indicate that TSMC’s capacity constraints have become severe enough that major AI chip designers are actively seeking alternatives. TSMC’s chief executive has publicly acknowledged that AI chip demand will exceed the company’s supply capacity for years, creating an opening for competitors like Intel to capture overflow demand.

Intel’s foundry business generated $4.5 billion in revenue in its most recent quarter, up modestly on a sequential basis but still operating at a loss as the company invests heavily in extreme ultraviolet lithography equipment and next-generation manufacturing processes. The company has positioned itself as the only foundry currently shipping chips using gate-all-around transistor technology at scale, marketed under the RibbonFET brand name.

Nvidia Testing Intel’s 18A Process

Separately, reports indicate Nvidia is evaluating whether Intel’s 18A manufacturing process can produce a forthcoming processor that combines four graphics processing units into a single package. The testing phase involves multi-project wafers, a standard industry practice for qualifying new manufacturing partners before committing to volume production.

Importantly, Nvidia has not yet placed a firm order with Intel. The graphics chip giant remains primarily reliant on TSMC for its AI accelerator production, and any shift would represent a major strategic realignment. However, the fact that Nvidia is conducting technical evaluations at all signals a meaningful change in the competitive landscape.

Intel’s advanced packaging capabilities, including its Embedded Multi-die Interconnect Bridge and Foveros technologies, are viewed as key differentiators that could help the company win business from customers seeking alternatives to TSMC’s constrained capacity. For US-based hyperscalers and government-linked workloads, Intel’s domestic manufacturing footprint offers supply chain diversification and reduced geopolitical risk compared to Taiwan-based production.

Analyst Perspectives

Last week, Mizuho raised its price target on Intel from $124 to $128 while maintaining a Neutral rating. The firm cited strong demand for so-called agentic AI applications, which require CPUs to orchestrate complex multi-step workflows rather than simply accelerating parallel computations. Mizuho noted that suppliers across the CPU ecosystem remain supply-constrained into 2027, indicating potential pricing power and high utilization rates.

Wells Fargo similarly increased its Intel price target from $85 to $110, keeping an Equal Weight rating. The firm’s recent channel checks during a Silicon Valley tour found robust AI data center build-out and growing demand for inference workloads. However, Wells Fargo warned that memory constraints appear likely to persist through 2027 and potentially into 2028, which could limit the pace at which system builders can absorb new CPU and accelerator capacity.

Both firms stopped short of upgrading Intel to Buy-equivalent ratings, suggesting that even after the recent rally, risk-reward remains balanced rather than overwhelmingly positive. The hesitation likely reflects ongoing execution risks around Intel’s 18A process node, which must deliver on promised performance, power efficiency, and manufacturing yields to convert current customer interest into sustained revenue growth.

The Agentic AI Catalyst

A broader theme emerging from recent semiconductor industry commentary is the rise of agentic AI workloads, which involve AI systems that plan, reason, and execute multi-step tasks autonomously. Unlike earlier generations of AI focused primarily on pattern recognition or content generation, agentic AI requires substantial CPU resources for orchestration, control flow, and data movement between different processing elements.

Both Intel and AMD have highlighted unexpectedly strong CPU demand tied to this shift. Industry forecasts suggest CPU requirements for AI workloads could increase three to four times by 2027 compared to current levels. For Intel, this represents an opportunity to reclaim strategic relevance in an AI landscape that initially appeared dominated by GPU-centric architectures.

However, the same supply constraints that are driving customers toward Intel also create execution risk. If Intel cannot bring new manufacturing capacity online smoothly, or if yields on advanced nodes disappoint, the company could struggle to capitalize on the demand window. Additionally, AMD and ARM-based server platforms are aggressively targeting the same agentic AI workloads, meaning Intel must win share rather than simply benefiting from rising tides.

Outlook

Capturing even a modest share of the AI accelerator manufacturing market could add billions in high-margin revenue by the end of the decade. Combined with a potential CPU renaissance driven by agentic AI workloads, Intel could return to sustained revenue growth and margin expansion after years of market share losses and profitability pressure.

The bear case centers on execution risk and the reality that much of the good news is back-end loaded. Nvidia has not committed to volume production, and even Google’s order lands primarily in 2028. TSMC is not standing still; the Taiwanese giant continues to invest heavily in capacity expansion and next-generation process technology. If TSMC resolves its capacity constraints faster than expected, or if AI capital expenditure growth slows, Intel’s window of opportunity could narrow quickly.

Markets are now pricing Intel as though the foundry turnaround will succeed, a sharp reversal from the skepticism that prevailed just months ago. Whether that optimism proves justified will depend on Intel’s ability to convert customer interest into profitable production at scale, a test that will play out over the next several quarters as 18A ramps and 2028 commitments draw closer to realization.