Google Expands Its Portfolio of Proprietary AI Chips
New processing units focus on data security and increased memory capacity for complex models.
Google has announced two new variants of its tensor processing units, known as TPUs. The TPU 8tx and TPU 8i Max models are part of the company's eighth generation of hardware and aim to expand cloud computing capabilities.
The TPU 8tx chip was developed in partnership with AMD. This component focuses on confidential model training to ensure that sensitive data remains protected and isolated even during active processing.
Meanwhile, the TPU 8i Max combines two inference chips into a single package. This architecture doubles memory capacity and bandwidth, allowing artificial intelligence models to process larger volumes of information during queries.
This move is part of a broader strategy to reduce reliance on external semiconductor suppliers. Technology companies and artificial intelligence laboratories are seeking to develop their own hardware to gain autonomy in the sector.
The initiative intensifies competition in the specialized semiconductor market. Google's technological advancement could pressure the global supply chain and alter cost dynamics for AI infrastructure development.
Spotted an error? Write to suporte@criptoico.com.br.
