Intel: Alphabet Orders 3M Custom AI Chips, Stock Jumps 11%

LONDON, Tuesday, June 9, 2026 — Intel shares surged 11.2% on Monday after The Information reported that Alphabet has placed an order with Intel to manufacture more than three million custom tensor processing units in 2028. Alphabet's Google has placed an order with Intel to manufacture more than three million tensor processing units in 2028, citing people with direct knowledge of the discussions. The deal is the first time Google's parent will use Intel as a contract manufacturer for its proprietary AI silicon, and it lands in the same week Nvidia used its Computex keynote to push past the data center and into the PC market.

Alphabet has selected Intel to manufacture more than 3,000,000 custom AI chips for its data centers. This is the first time Alphabet will use Intel as a contract manufacturer for its proprietary AI hardware. The move responds to tight capacity at Taiwan Semiconductor and growing demand for AI computing power.

Key Takeaways

• Intel jumped 11.2% on June 8 on news of an Alphabet TPU manufacturing order exceeding three million units by 2028, its biggest single-day move in months.
• The order breaks TSMC's effective monopoly on advanced AI accelerator production for a top-three hyperscaler.
• Nvidia's RTX Spark, announced May 31 at Computex, brings 1 petaflop of AI performance to Windows laptops and pulled AMD, Intel and Qualcomm shares lower on the news.
• Alphabet is funding the buildout with an upsized $84.75 billion equity raise priced June 2, including a $10 billion Berkshire Hathaway private placement.
• Intel also used Computex to launch Xeon 6+ on its 18A node and a rack-scale AI infrastructure partnership with Foxconn.

Context & Analysis

The Alphabet-Intel order matters less for its 2028 timing than for what it signals today. The news is the latest sign that a turnaround at Intel is gathering steam after years of management blunders allowed TSMC to take the lead in chip manufacturing. Google designs its TPUs in-house and has historically relied on TSMC for fabrication. Splitting that work introduces a second source for capacity that has become the binding constraint on hyperscaler AI roadmaps.

The order also arrives alongside parallel evaluation by Nvidia. Nvidia has not placed an order with Intel yet, but has been evaluating whether the company's technology can be used to make a processor that combines four graphics chips into a single unit, the report said. Tesla CEO Elon Musk has separately committed to Intel's 14A process for the Terafab project in Austin, per the same Reuters report. Three of the largest AI silicon buyers are now publicly engaged with Intel Foundry.

The financing side is equally aggressive. Alphabet announced the pricing of registered public offerings of Class A Common Stock, Class C Capital Stock and depositary shares representing interests in mandatory convertible preferred stock. The gross proceeds, together with a $40 billion at-the-market offering program and a concurrent $10 billion private placement, represent a total equity raise of $84.75 billion. The equity capital raise was upsized from the previously announced total equity raise of $80 billion.

Related: Nvidia Signals 2026–2030 AI Chips Demand Surge As Investors Reposition

Related: AI Chips Talent Reset: Nvidia, TSMC and AWS Rework Roles as Packaging and In-House Silicon Surge

Competitive Landscape

Nvidia did not sit still. At Computex 2026 in Taipei on May 31, Jensen Huang unveiled the RTX Spark superchip for Windows on Arm. Huang announced NVIDIA RTX Spark, which brings 1 petaflop of AI performance to slim Windows laptops and compact desktops. Built with MediaTek and running Microsoft Windows, it powers the first PCs purpose-built for personal agents — always on, always local. Huang described RTX Spark as "everything we've learned over 33 years distilled into one chip," pairing an NVIDIA Blackwell RTX GPU with 6,144 CUDA cores and 1 petaflop of AI performance with a custom 20-core Grace CPU "built in partnership with MediaTek, fused by NVLink." Independent research organizations have documented comparable patterns. As highlighted in annual shareholder communications, that market conditions support continued investment.

For deeper context, see our AI Chips analysis: "AI Chips Interoperability Breakthroughs: UCIe Compliance, ONNX Runtime Upgrades, Cloud Cross-Accel Pilots".

The spec sheet is aggressive. At full strength, this chip offers up to 20 Arm CPU cores, a Blackwell GPU with 6,144 CUDA cores, 128GB of LPDDR5X RAM, and up to 300 GB/s of memory bandwidth. That powerful CPU and GPU, connected over NVLink C2C, and the large memory pool give AI agents and 120-billion-parameter models plenty of power and space for long-running tasks with context lengths stretching to a million tokens, according to Nvidia. Wall Street read it as a direct attack on AMD, Intel and Qualcomm in the PC segment. Nvidia's announced entry into the PC chip market sent shares of AMD, Intel and Qualcomm lower on Monday as Wall Street recognized the threat.

Microsoft sits on both sides. Nvidia says Microsoft will reveal more details of this agentic AI transformation at its upcoming Build conference. RTX Spark systems will begin arriving in the fall of 2026. Microsoft also unveiled Majorana 2 on June 2, a separate quantum bet. Microsoft unveils Majorana 2, its next-generation topological quantum chip developed with the help of Microsoft Discovery's agentic AI. Majorana 2's new features include a new materials stack enabling a 1,000-fold improvement in reliability over the prior generation of qubits, with a mean qubit lifetime of 20 seconds and instances lasting as long as one minute. Microsoft now expects to achieve a scalable quantum computer by 2029, cutting its original timeline in half.

Additional coverage: How ASML and TSMC Will Impact Global AI Chips Market and AI Supply Chain in 2026

Additional coverage: AMD Unveils MI400 AI Chip Series With 432GB HBM4 Memory at CES 2026

What It Means

For Enterprise Buyers

Capacity is the new constraint, not architecture. Hyperscaler willingness to second-source advanced AI silicon at Intel reduces tail risk for cloud customers dependent on TPU- and GPU-backed services. Buyers planning multi-year inference contracts should pressure cloud vendors for explicit supply-chain diversification language. On the device side, RTX Spark moves a credible 120B-parameter inference workload onto local hardware, changing the build-versus-buy math for regulated industries with data-residency rules.

Related: Amazon, OpenAI & Apple Signal AI Chip Shift with Trainium in 2026

For Investors

Intel's 11% move repriced Foundry optionality, not the legacy x86 business. One Alphabet order does not resolve Intel's structural challenges, but it does signal that hyperscalers are willing to diversify their AI silicon supply chains beyond a single dominant vendor. The same dynamic supports Broadcom, AMD's MI-series roadmap, and any TPU-adjacent ASIC supplier. The risk: 2028 is far enough out that yield slippage on Intel 18A — or its successor 14A — could unwind the trade.

For deeper context, see our related analysis: "AI chips draw record capital as hyperscalers and fabs reset the stack."

For deeper context, see our Blockchain analysis: "Blockchain Investment Market Trends: Institutional Inflows And Tokenization Push".

Forward Outlook

Three milestones drive the next leg. First, Microsoft Build will detail the agentic Windows stack and how Surface Laptop Ultra ships against the RTX Spark roadmap in fall 2026. Second, Intel needs to demonstrate 18A yield on the Xeon 6+ ramp before customers commit to follow-on Alphabet-style orders. Third, Alphabet's $84.75 billion raise will start showing up as data-center capex in the second half, putting concrete spend behind the Intel relationship. Watch Q3 earnings prints from both companies for confirmation.

Related: Top 7 AI Chips Priorities Hyperscalers Accelerate for 2026

Additional coverage: AI chip startups surge: funding spikes, new architectures, and a supply chain squeeze

For deeper context, see our related analysis: "UCIe Consortium Launches Chiplet Certification As NIST Proposes Secure AI Compute Baseline."

Sources include company disclosures, regulatory filings, analyst reports, and industry briefings.

Related Coverage

• More AI Chips Analysis
• Latest Technology News