Intel (INTC) locked in a multiyear collaboration with Google committing the search giant to multiple generations of Xeon processors for AI training, inference, and general-purpose computing, plus expanded co-development of custom infrastructure processing units (IPUs) that offload networking and storage tasks. AMD (AMD) posted 34% overall revenue growth to $34.6B in fiscal 2025 driven by EPYC CPUs, while Nvidia (NVDA) maintains the GPU market with a trailing P/E around 37.

Agentic AI workloads are shifting the computing bottleneck back toward general-purpose CPUs, validating Intel’s strategy that balanced systems combining CPUs and accelerators are essential to AI infrastructure scale.

The analyst who called NVIDIA in 2010 just named his top 10 AI stocks. Get them here FREE.

AI workloads have stretched data-center capacity to the limit. Hyperscalers now spend hundreds of billions annually to keep pace, yet GPUs handle only part of the job. CPUs still orchestrate training runs, manage inference at scale, and deliver the balanced systems that prevent bottlenecks.

Today, Intel (NASDAQ:INTC) locked in another piece of that puzzle. The company announced a multiyear collaboration with Google that commits the search giant to multiple generations of Xeon processors across its AI infrastructure. The stock is up more than 3% in morning trading to almost $61 a share..

According to Intel's press release, Google Cloud will continue deploying Xeon processors for AI training coordination, latency-sensitive inference, and general-purpose computing. The pact specifically includes the latest Xeon 6 chips powering C4 and N4 instances, plus expanded co-development of custom ASIC-based infrastructure processing units (IPUs). These IPUs offload networking, storage, and security tasks from the host CPU, freeing cycles for actual compute.

READ: The analyst who called NVIDIA in 2010 just named his top 10 AI stocks

Intel CEO Lip-Bu Tan put it plainly: “Scaling AI requires more than accelerators -- it requires balanced systems. CPUs and IPUs are central to delivering the performance, efficiency, and flexibility modern AI workloads demand.” Google’s Amin Vahdat, SVP and chief technologist for AI infrastructure, added that Intel’s Xeon roadmap “gives us confidence that we can continue to meet the growing performance and efficiency demands of our workloads.”

The collaboration builds on nearly two decades of partnership and targets better energy efficiency plus lower total cost of ownership across Google’s global fleet. No dollar value was disclosed, but the multi-generation commitment removes guesswork for Intel’s factory ramp.

intc

Let’s compare the scoreboard. Nvidia (NASDAQ:NVDA) still owns the GPU crown, with trailing P/E around 37 and data-center revenue that dwarfs everyone else. Advanced Micro Devices (NASDAQ:AMD) posted 34% overall revenue growth to $34.6 billion in fiscal 2025, driven by EPYC CPUs and Instinct GPUs. Intel’s Data Center and AI group, however, grew 5% to $16.9 billion for the full year.

That 5% gain looks modest until you remember the context: Intel supplies the x86 foundation that runs 80% of the world’s servers. Xeon 6 -- built on the new 18A process node -- slots directly into heterogeneous AI racks where GPUs handle raw matrix math and CPUs manage orchestration.

The Google deal underscores a simple truth -- agentic AI workloads are shifting the bottleneck back toward general-purpose compute. Believe it or not, the same Xeon family that powered Google’s earliest racks now gets a fresh AI mandate.

Intel’s fiscal 2025 results show clear progress. Revenue held flat at $52.9 billion, but non-GAAP gross margin expanded 70 basis points to 36.7% and non-GAAP operating margin swung from negative 0.5% to positive 5.5%. Non-GAAP EPS turned positive at $0.42, up from a $0.13 loss the prior year. Operating cash flow reached $9.7 billion, though levered free cash flow remained negative at roughly $4.9 billion amid heavy 18A investments.

At today’s price, the stock trades at a forward P/E of 101, a price-to-sales ratio of 5.78, and price-to-book of 2.58. Intel's trailing P/E remains negative because of prior losses, but the forward multiple reflects analyst bets on margin recovery. No dividend is currently paid -- the payout was suspended in 2024, but supply constraints that capped Q4 revenue should ease after the first quarter when management expects the trough before sequential improvement.

In short, the Google pact validates Intel’s bet that CPUs remain essential to AI scale. It is clear that a hyperscaler committing across multiple Xeon generations is tangible proof of demand. That said, the foundry business still bleeds cash and the valuation leaves little room for missteps. Recent wins in its foundry business, though, promise to change that trajectory.

Intel remains a hold for existing shareholders, who will watch the April 23 earnings report for confirmation that supply is ramping up and margins are holding. If those boxes get checked, the multi-year tailwind from deals like this one with Google could deliver steady gains -- all without the drama of pure GPU plays. Intel isn’t done winning; it just needs to execute.

Wall Street is pouring billions into AI, but most investors are buying the wrong stocks. The analyst who first identified NVIDIA as a buy back in 2010 — before its 28,000% run — has just pinpointed 10 new AI companies he believes could deliver outsized returns from here. One dominates a $100 billion equipment market. Another is solving the single biggest bottleneck holding back AI data centers. A third is a pure-play on an optical networking market set to quadruple. Most investors haven't heard of half these names. Get the free list of all 10 stocks here.