Google Cloud expands custom silicon family with TPU 8tx and TPU 8i Max
Google Cloud has announced two additions to its eighth-generation custom accelerator family aimed at secure enterprise workloads and higher-capacity artificial intelligence inference.
FTMQ SI, written by our newsroom2 views

Google Cloud announced two new accelerators in its eighth-generation Tensor Processing Unit family on Monday. Data Center Knowledge, MarkTechPost, and Pulse 2.0 reported that the company unveiled the TPU 8tx for secure artificial intelligence workloads and the TPU 8i Max for high-capacity inference. The processors address demand for confidential processing and agentic applications requiring large context windows. [1][2][3]
The TPU 8i Max combines two TPU 8i chips into a single package, according to Data Center Knowledge, MarkTechPost, and Wccftech. Data Center Knowledge and Pulse 2.0 reported that this dual-chip design doubles compute, static random-access memory, and memory bandwidth compared to the standard TPU 8i processor. [1][2][3][5]
The TPU 8tx variant introduces hardware-based confidential inference, according to MarkTechPost. Wccftech reported that Google is pairing the TPU 8tx with custom AMD EPYC central processing unit hosts. The Next Platform noted that the new chips are variants of the Sunfish TPU 8t and Zebrafish TPU 8i accelerators that Google initially revealed in April. [2][4][5]
Tensor Processing Units are application-specific integrated circuits developed by Google for neural network machine learning, supporting frameworks such as TensorFlow, Jax, and PyTorch. Google began using the processors internally in 2015 before making them available for third-party cloud infrastructure use in 2018. [10]
The Next Platform reported that Google will roll out the new accelerators next year. Bloomberg reported that availability will begin in 2027, according to MarkTechPost. [2][4]
In short
- Google Cloud announced the TPU 8tx and TPU 8i Max processors for enterprise artificial intelligence.
- The TPU 8i Max combines two TPU 8i chips in a single package to double compute and memory bandwidth.
- The TPU 8tx features hardware confidential inference and pairs with custom AMD EPYC host processors.
Sources
Every paragraph above points to the numbered items it rests on. Read the originals here.
- [1]Google Adds TPU 8tx, 8i Max for Enterprise AIData Center Knowledge, 8d ago (the report this story comes from)
- [2]Google Unveils TPU 8i Max and TPU 8tx: A Dual-Chip TPU and a Confidential TPUMarkTechPost, 8d ago
- [3]Google Unveils TPU 8tx And TPU 8i Max With 2X Compute And Enhanced Security For AI AgentsPulse 2.0, 8d ago
- [4]Google Rejiggers TPU Accelerators For Broader AI Inference WorkThe Next Platform, 8d ago
- [5]Google Pairs Its New TPU 8tx With Custom AMD EPYC CPU Hosts, While The TPU 8i Max Combines Two TPU 8i Chips In A Single Package To Offer Phenomenal Compute And Memory DensityWccftech, 8d ago
- [6]Google advances agentic AI custom silicon expanding TPU family with 8tx and 8i MaxSiliconANGLE, 8d ago
- [7]Google adds new TPU 8tx and TPU 8i Max for AI workloadsTestingCatalog AI News, 8d ago
- [8]Google's AI chip business nears 100 customers as it adds TPU 8tx and 8i MaxPPC Land, 8d ago
- [9]Alphabet Stock Rises as Google Adds Two New TPU VariantsYahoo Finance, 8d ago
Background
- [10]Tensor Processing Unit on Wikipedia
- [11]Mistral AI on Wikipedia
Our newsroom writes these reports with the help of software, from the 11 sources listed and nothing else, and checks them against those sources. Facts can still be wrong or move on; the originals are the record. Spotted a mistake? Write to daniel@monsterkong.com.
Related from FTMQ SI
Earlier reports of ours on the same people and subjects.
- Google announces TPU 8tx and TPU 8i Max acceleratorsModels & Products, 8d ago
- Microsoft unveils Maia 300 custom silicon for cloud AI inference workloadsTop Stories, 36d ago
- Oracle expands cloud hardware with Google TPUs and OpenAI inference chipsModels & Products, 21d ago
- Google Expands TPU Inference Ecosystem to Challenge Nvidia Software DominanceIndustry Labs, 12d ago
- Dell and AMD Consolidate Enterprise AI Hardware Into Rack Scale SystemsModels & Products, 15d ago
More in Chips & Infrastructure
- VOLTA infostealer malware uses Claude AI to rank stolen victim data16d ago
- Advanced Photonics Coalition releases standards to accelerate optical component manufacturing23d ago
- Researchers test platinum tungsten alloy for thinner extreme ultraviolet mask absorbers23d ago
- Proton hosts public surveillance map identifying over 300000 Flock cameras24d ago
- Developer uses AI agents to rebuild Call of Duty game source code36d ago
- Semiconductor startups raised 8.5 billion dollars in third quarter 202636d ago
Get the day in one email
Reports like this one, the top news of the last 24 hours, every morning. Free, one email a day; readers can comment under every report.
By signing up you agree to our terms and privacy policy. Unsubscribe any time.

Comments
Loading
Join the conversation
Comments are open to readers of our daily email: the top news of the last 24 hours, every morning, free. Sign up and the comment box opens.
Already on the list? Enter the same address and we will send a sign-in link.
By signing up you agree to our terms and privacy policy. Unsubscribe any time.