Google Orders 3 Million AI Chips From Intel for 2028
💡Google's massive 3M chip order from Intel signals a major shift in AI infrastructure supply chain strategy.
⚡ 30-Second TL;DR
What Changed
Google plans to source over 3 million specialized AI chips from Intel.
Why It Matters
This partnership signals a shift in the AI hardware landscape, potentially reducing reliance on single-source suppliers like TSMC. It strengthens Intel's foundry business as a viable competitor for high-end AI silicon.
What To Do Next
Monitor Intel's foundry roadmap and process node advancements to evaluate if your future infrastructure needs can leverage their custom silicon services.
Key Points
- •Google plans to source over 3 million specialized AI chips from Intel.
- •The production timeline for these chips is set for 2028.
- •The deal underscores Intel's growing role as a foundry partner for major tech firms.
🧠 Deep Insight
Background and context from public sources — not the original article. 11 sources cited.
🔑 Enhanced Key Takeaways
- •The chips Google has ordered are specifically its in-house Tensor Processing Units (TPUs), which are custom AI accelerators designed for training and running AI models.
- •This deal represents a significant win for Intel Foundry Services (IFS), positioning it as a stronger competitor to Taiwan Semiconductor Manufacturing Company (TSMC) amidst TSMC's reported capacity constraints and advanced packaging squeeze.
- •The order follows months of Google testing Intel's advanced packaging technology, and Intel is also in discussions with Amazon and has Nvidia evaluating its 18A manufacturing process and advanced packaging for future GPU architectures, indicating a broader industry trend towards diversifying chip supply chains.
- •Intel's advanced packaging technologies, such as Embedded Multi-die Interconnect Bridge (EMIB), are highlighted as a key differentiator that could enable Intel to quickly gain market share in AI chip manufacturing by offering cost-effective, flexible, and scalable solutions.
- •Google has already begun selling TPU computing capacity to other companies like Apple and Meta Platforms, and Morgan Stanley estimates Google could produce over six million TPUs across 2027 and 2028.
🛠️ Technical Deep Dive
- The chips ordered are Google's Tensor Processing Units (TPUs), which are custom AI accelerators.
- Google's latest generation of TPUs, the 8th generation (TPU 8t and 8i), were unveiled at Cloud Next 2026, with the 8t variant focused on training and the 8i optimized for inference.
- These 8th generation TPUs are designed for scalability, capable of forming clusters exceeding one million units.
- Future TPU variants may leverage Intel's Embedded Multi-die Interconnect Bridge (EMIB) technology, which facilitates high-speed connections between multiple chiplets within a package without requiring a large silicon interposer, offering benefits in cost, flexibility, and scalability.
- Intel's advanced manufacturing processes, specifically its cutting-edge 18A (1.8-nanometer-class) process, are being evaluated by other major AI chip designers like Nvidia for their next-generation GPU architectures.
🔮 Future ImplicationsAI analysis grounded in cited sources
⏳ Timeline
📎 Sources (11)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
Weekly AI Recap
Read this week's curated digest of top AI events →
👉Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: Bloomberg Technology ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
The weekly digest
One email a week. Unsubscribe anytime.