Google Introduces Two New AI Systems for Enterprises
Promoted
Google has expanded its eighth-generation Tensor Processing Unit (TPU) portfolio with two new AI computing systems, TPU 8tx and TPU 8i Max, designed to support secure enterprise AI applications and complex workloads involving large volumes of data.
Advertisement
Announced on November 9, 2026, the systems address two growing requirements in enterprise AI: protecting sensitive information during processing and providing greater computing and memory capacity for advanced AI models and autonomous agents.
Google said the additions take its eighth-generation TPU portfolio to four systems, following the introduction of TPU 8t and TPU 8i in April 2026.
TPU 8tx Focuses on Secure Enterprise AI
The TPU 8tx is designed for government organisations and businesses in regulated sectors that need to process confidential information using AI.
Advertisement
It is built on the same silicon as Google’s TPU 8t and combines it with fifth-generation AMD EPYC CPU hosts and hardware-level security features. The system is designed to protect sensitive data and AI model weights even while they are being processed.
According to Google, TPU 8tx can deliver up to twice the speed of TPU 8t in terms of latency for certain workloads, with almost no performance penalty from its security features.
The system is intended to support:
- Secure AI inference using confidential business data.
- Fine-tuning AI models using private information.
- Training AI agents through reinforcement learning.
- AI applications for government and regulated industries.
Advertisement
TPU 8i Max Targets Complex AI Workloads
The second system, TPU 8i Max, is designed for AI applications that require large context windows and higher memory capacity.
Google has combined two TPU 8i chips into a single package. According to the company, this configuration provides twice the computing capacity, twice the on-chip static random-access memory (SRAM) and twice the high-bandwidth memory (HBM) bandwidth of the standard TPU 8i.
The system is designed to support applications where AI models need to process large amounts of information and handle multiple tasks simultaneously.
Potential use cases include:
- AI-powered coding assistants working across large software projects.
- Mathematical reasoning and complex problem-solving.
- Drug discovery and scientific research.
- Enterprise AI agents processing extensive documents and data.
Google also said the system can provide greater computing capacity within the existing rack footprint, helping data centres increase computing density.
Google Expands Its Enterprise AI Infrastructure Business
The announcement reflects Google’s efforts to expand its AI infrastructure business as organisations move towards deploying AI agents and advanced AI applications.
Google said nearly 100 customers, including research organisations, trading firms, retailers and universities, are using its TPU systems.
The company is also working to make its TPUs more accessible to developers through support for PyTorch and open-source AI software. This is expected to reduce the technical barriers for organisations looking to use Google’s custom AI processors.
Google Cloud reported revenue of $30.71 billion in the September 2026 quarter, more than double the year-earlier figure. Google also said TPU sales made a meaningful contribution to its quarterly results.
The latest additions underline the growing importance of specialised computing infrastructure in enterprise AI. As businesses deploy more AI agents, demand is increasing for systems that can handle complex processing requirements while maintaining data security.
TPU 8tx and TPU 8i Max are new system configurations based on Google’s existing TPU 8t and TPU 8i silicon, rather than two entirely new chip architectures.