7/23/2026, 1:05:30 PM · infrastructure

Google Splits TPU Lineage Into Dedicated Training and Inference Chips, Claims Structural Cost Advantage in AI Serving

At Google Cloud Next 2026, Google unveiled the TPU 8t and TPU 8i — the first time in its Tensor Processing Unit program's history that training and inference workloads have received entirely separate silicon — while its flagship Gemini 3.5 Pro model remains delayed over coding performance shortfalls.

Background: A Decade of Custom Silicon Reaches an Inflection Point

Google's Tensor Processing Unit (TPU) program has run for roughly ten years as a unified accelerator serving both model training and inference. <cite index="26-1">At Cloud Next 2026 on April 22, Google split TPU 8 into TPU 8t (training) and TPU 8i (inference) — the first time the TPU program has separated training and inference silicon in its ten-year history.</cite> The architectural bifurcation marks a deliberate bet that the economics of AI are now dominated by serving rather than by building models.

The Two Chips

<cite index="21-1">The TPU 8t was engineered for large-scale training and designed with Broadcom, while the TPU 8i was engineered for reasoning and serving and designed with MediaTek.</cite> The division of labor is intended to eliminate the performance compromises inherent in general-purpose Application-Specific Integrated Circuits (ASICs).

<cite index="19-3,19-4,19-5">The TPU 8i, codenamed Zebrafish, is the more technically distinctive of the two chips; rather than chase training flops, it is engineered around latency — the metric that determines whether an AI agent feels real-time or sluggish. Each TPU 8i delivers 10.1 FP4 petaflops of compute, fed by 384 MB of on-chip SRAM (triple the prior generation), 288 GB of HBM, and 8.6 TB/s of memory bandwidth.</cite>

<cite index="22-7,22-8">The large SRAM enables the active working set of a large model to remain entirely on-chip during inference, eliminating the latency of loading weights from HBM on every token generation. The chip uses a new Collective Acceleration Engine that offloads distributed inference communications — the all-reduce and all-gather operations that dominate synchronization overhead in large multi-chip inference clusters.</cite>

Performance-Per-Dollar Claims

<cite index="27-1">According to Google's own Cloud Blog technical deep dive, TPU 8i delivers up to 80% performance-per-dollar improvement over Ironwood TPU, particularly at low-latency targets for large Mixture-of-Experts (MoE) models.</cite> When stacked against earlier generations, <cite index="21-5">Google is mapping to roughly 6x more compute per unit of electricity versus five years ago, aiming at up to 2x performance-per-watt over Ironwood.</cite>

This efficiency trajectory compounds gains Google had already recorded internally. <cite index="4-1">On its 2025 earnings call, Google said it had cut Gemini's per-unit serving cost by roughly 78% that year, crediting model optimization plus efficiency and utilization gains from its own hardware.</cite> <cite index="20-13,20-14">Gemini is served exclusively on TPUs, while competing models such as Claude and GPT-5.5 are not, meaning the fundamentally different architecture that Gemini runs on is a key driver of Google's lower cost per token.</cite>

Ecosystem and Anchor Customers

<cite index="26-4">Anthropic anchors both generations with a 3.5 gigawatt capacity deal for 2027; Google simultaneously inked a Vera Rubin partnership with NVIDIA (Graphics Processing Unit), framing the play as a multi-silicon "AI Hypercomputer" rather than a head-on replacement strategy.</cite> <cite index="22-10">Both TPU 8t and 8i will be generally available on Google Cloud Platform later in 2026, accessible as standalone instances or as part of Google's AI Hypercomputer full-stack platform.</cite>

Gemini 3.5 Pro Delay

The infrastructure announcement arrives against a backdrop of model-level turbulence. <cite index="11-1">Google launched Gemini 3 in late 2025, and the 3.5 generation was supposed to continue a rapid release cadence, with the Flash version announced in May and the Pro version promised a month later.</cite> <cite index="12-4">Google's most powerful flagship AI model, Gemini 3.5 Pro, has fallen months behind schedule as engineers work to improve its coding capabilities, according to Bloomberg.</cite> <cite index="11-4">Google updated the data used to train Gemini in an attempt to improve its coding skills, and the results were disappointing.</cite>

<cite index="14-11">Providing comment on the matter to Bloomberg, Google said: "We're shipping quickly across a wide range of models while keeping them highly cost-effective for customers."</cite> <cite index="10-8">The report hit investor sentiment immediately, with Alphabet shares falling about 4% after the news rolled out.</cite> <cite index="11-5">Google has not announced a new launch date.</cite>

Structural Significance

<cite index="5-5,5-6">The TPU 8i chip is part of a broader infrastructure push that extends Google's decade-long custom silicon program into a new phase explicitly designed around the inference economics of Large Language Models (LLMs) rather than training alone — because inference is now the dominant cost driver across frontier AI.</cite> <cite index="20-3,20-4,20-5">According to industry analysts, inference could take the larger share of compute demand as soon as 2027; McKinsey estimates that in 2026, training and inference will each account for roughly 31.2 gigawatts of data center demand, but by 2027, inference becomes the larger share.</cite> Whether the silicon advantage translates into a durable competitive moat will depend in part on whether Google can close the model-capability gap that the Gemini 3.5 Pro delay has exposed.

Cross-references

Sources

  1. [1]
    Google TPU v8 vs Nvidia: How Inference Is Rewriting the AI Market
  2. [2]
    AI Inference Costs 2025: Why Google TPUs Beat Nvidia GPUs by 4x
  3. [3]
    Google TPU vs NVIDIA GPU | Introl Blog
  4. [4]
    In-House TPUs: Just How Deep Is Google's Chip-Cost Moat? | Penchan
  5. [5]
    Google's In-House AI Chip Strategy and What It Means for the Gemini Cost Structure - FourWeekMBA
  6. [6]
    Google TPU 8i Cost Analysis: 80% Better Performance-Per-Dollar Explained - Artificial Intelligence World
  7. [7]
    Google TPUv7: The 900lb Gorilla In the Room
  8. [8]
    The chip made for the AI inference era – the Google TPU
  9. [9]
    Why TPU Can Cut Costs by 80% at Scale & Google's Balance Sheet Strategy Hints Aggressive Market Move
  10. [10]
    Google Delayed Gemini 3.5 Pro Because Its Coding Fell Short
  11. [11]
    Alphabet's Gemini 3.5 Pro Is Late and the Stock Is Slipping. Is the AI Leader Falling Behind?
  12. [12]
    Google's Gemini 3.5 Pro release faces delay: report | Seeking Alpha
  13. [13]
    Google Delays Gemini 3.5 Pro Over Coding Issues: Report
  14. [14]
    Google Gemini 3.5 Pro faces "delays" over coding performance misses - Neowin
  15. [15]
    Google delays Gemini 3.5 Pro to boost its AI model
  16. [16]
    www.mexc.com
  17. [17]
    next BIG future
  18. [18]
    DEV Community
  19. [19]
    Google TPU 8t and 8i: 121 Exaflops, $21B Nvidia Challenge
  20. [20]
    Google Cloud Next 2026: Google Cloud Bifurcates the AI Future – Specialized TPU 8t and 8i Architectures Signal the End of General-Purpose Silicon
  21. [21]
    Google TPU 8 Is Here. It Is Not Replacing NVIDIA. That Is the Whole Point.
  22. [22]
    Google launches TPU 8 chips to speed AI training and cut costs
  23. [23]
    Google launches Ironwood TPU and previews eighth-gen split into training and inference chips at TSMC 2nm
  24. [24]
    Google TPU 8: 9,600-Chip Pod, 80% Cheaper Inference
  25. [25]
    TPU 8t and TPU 8i technical deep dive | Google Cloud Blog