Radar / AI infrastructure / Google TPU 8tx and TPU 8i Max
Google TPU 8tx and TPU 8i Max: confidential inference and double-chip long context
Google's eighth-gen TPU family grows to four systems in one year: 8tx for confidential inference, 8i Max for massive context windows.
Why it matters
TPU 8tx is the first cloud TPU with hardware-enforced confidential computing via AMD EPYC hosts, and TPU 8i Max packs two 8i chips on one package for double the compute, SRAM and HBM bandwidth.
What you could build with it
A cloud team serving regulated customers could prototype confidential fine-tuning on TPU 8tx once it ships in 2027, checking whether hardware-enforced isolation removes the compliance review that currently blocks agent deployments on sensitive data.
Does it hold up?
Too early to judge: announced Nov 9 with availability in 2027, no pricing or regions yet; the systems are a menu change, not a bill, until they ship.
Built with Google TPU 8tx and TPU 8i Max
- Google adds TPU 8tx and TPU 8i Max to its eighth-generation TPU lineglonce · Independent read of the announcement, including the TorchTPU open-source note and the four-systems-in-one-year first.
- Google's New TPU 8tx Keeps AI Data Encrypted While It Runswebpronews · AMD's own announcement confirms 5th Gen EPYC hosts and first hardware-enforced confidential computing on a cloud TPU.
- Google TPU 8tx and 8i Max: two new AI chips to take on Nvidiamadrobot · Specs and gaps: no prices or availability dates in Google's post, and performance figures are Google's own.
Learn more
First spotted on article: source.
More AI infrastructure
vLLM v0.34.0: extensible KV cache by default, Model Runner V2 only, and 10 new modelsThe open inference engine's latest release spans 1,061 commits from 412 contributors, adds 10 models including Mistral…infra · JEV 0.86llama.cpp v1.0.0: the local-AI engine's first stable releasellama.cpp and ggml hit v1.0.0, promising semantic versioning so breaking C API changes only arrive with major bumps.infra · JEV 0.84colibri: run 744B-parameter MoE models on consumer hardwarePure-C inference engine with zero dependencies that streams MoE experts from NVMe disk, keeping hot experts on GPU…infra · JEV 0.78claude-mem: persistent memory plugin that lets coding agents pick up where the last session left offPlugin that hooks five Claude Code lifecycle events, compresses session activity into a local SQLite + Chroma store…infra · JEV 0.74FreeToken: datacenter-scale MoE serving on desktop GPUsAn edge-native mixture-of-experts serving engine from UC Berkeley researchers that runs 290B+ parameter MoE models on…infra · JEV 0.74GBrain: Garry Tan's self-wiring memory layer for AI agentsY Combinator CEO Garry Tan's open-source, markdown-first knowledge layer that gives OpenClaw and Hermes agents…infra · JEV 0.71
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.