پوشش زنده --:--:-- Tuesday, 10. November 2026 خبرگزاری خلیج نیوز
خبرگزاری خلیج نیوز صدای خلیج فارسkhalignews.com
Google Expands Eighth Generation Tensor Processing Hardware with TPU 8tx and 8i Max
Technology

Google Expands Eighth Generation Tensor Processing Hardware with TPU 8tx and 8i Max

منبع تصویر: timesofoman.com

By Khalij News Agency Editorial 2 min Read time 0

Google has introduced two new tensor processing systems, namely TPU 8tx and TPU 8i Max, thereby expanding its eighth generation of accelerators to meet the emergent needs of businesses in the field of artificial intelligence (AI). This development marks the first time the company has introduced four TPU systems in a single calendar year.

New Systems and Their Performance

This announcement comes after the unveiling of the 8t and 8i processors earlier this year. Google stated that the latest additions are designed to target specific infrastructure bottlenecks as businesses transition from human computations to automated operations.

In a statement, Google said, "Today, we are taking a step forward and expanding our eighth generation TPU lineup with two new systems: TPU 8i Max for wide context windows and the highest computational density per rack, and TPU 8tx for secure and confidential inference." The company also added, "This is the first time we have introduced four TPU systems in a year." According to Google, this division now serves nearly 100 external customers, including commercial enterprises, advanced research centers, universities, and commercial retailers.

Development and Design of the Systems

The statement emphasized that this announcement focuses more on complete computing systems rather than raw silicon design. While the physical chip manufacturing is limited to multi-year production plans, the system-level design combines computation, optical networks, memory, and storage configurations around modular silicon blocks to deliver specialized hardware in shorter timeframes.

Google noted that through internal design and integration into the design process, compiler engineers and network specialists have been coordinated with researchers at Google DeepMind. The company uses AI systems, including AlphaChip and Gemini models, to automate and deliver physical chip designs within hours, supporting the ongoing development of accelerators in its cloud computing infrastructure.

Source: timesofoman.com