Alphabet shares (GOOGL) rose on Monday after Google introduced two new versions of its in-house artificial intelligence chips.
The company is expanding its eighth-generation Tensor Processing Unit (TPU) portfolio as it looks to enhance its position in the AI hardware market.
The new systems, TPU 8tx and TPU 8i Max, are designed to address different computing requirements.
This includes secure AI processing and the growing memory demands of advanced AI models.
The announcement comes as Google steps up efforts to position its TPUs as an alternative to Nvidia’s graphics processing units (GPUs).
The move also coincides with increased scrutiny of Nvidia following reports of an investigation into an alleged foreign intelligence effort targeting its chief executive, Jensen Huang.
Alphabet shares rose 1.93% to $290.27 on Monday, while Nvidia stock fell around 5.3% to $176.75.
Google targets secure AI computing with TPU 8tx
Google said the TPU 8tx is designed for businesses and government agencies that need to run AI applications using sensitive information.
The system is optimised for secure and confidential inference.
It will allow organisations to fine-tune models and train AI agents using private data without moving that information between separate computing clusters.
According to Google, the hardware-level security features are intended to protect both private data and AI models during processing.
The system also delivers up to twice the lower latency compared with TPU 8t, supported by a faster network designed for AI reasoning workloads.
The TPU 8tx builds on Google’s existing eighth-generation chip architecture, which was developed in collaboration with Broadcom.
The company initially introduced the TPU 8 lineup for both AI model training and inference.
Google said its TPU business now serves nearly 100 customers, including AI research laboratories, trading firms and retailers.
TPU 8i Max addresses AI memory bottlenecks
The second system, TPU 8i Max, is designed to address the so-called memory wall, a bottleneck that occurs when processors can perform calculations faster than memory systems can supply the required data.
Google said the system combines two TPU 8i chips into a single package, doubling computing power, SRAM capacity and high-bandwidth memory bandwidth compared with TPU 8i.
The configuration is intended to support increasingly complex AI workloads, including agents that process large documents, software codebases and lengthy videos.
“TPU 8i Max scales compute and memory in lockstep…so processors remain equipped with essential data and so that large workloads do not stall,” Google said.
The company also said the configuration allows twice the silicon to fit into the same rack space, potentially increasing computing density without requiring additional physical data-centre space.
Google expands software access as Nvidia competition intensifies
Google is also working to make its chips more accessible to developers, addressing one of Nvidia’s key advantages: the widespread adoption of its CUDA software platform.
In October, Google open-sourced TorchTPU, allowing developers to run existing PyTorch code and Hugging Face models on its accelerators without rewriting their applications.
The company also introduced dynamic slicing technology, which it said can reduce computing job startup times by up to five times and accelerate recovery from hardware failures by up to 4.5 times.
Google has also announced that its TPUs will become available through Oracle’s cloud infrastructure platform from 2027.
This will mark the first time the chips are offered through a hyperscale cloud platform outside Google’s own infrastructure.
“Our aim is to make TPUs more easily accessible to every developer and Cloud customer,” Google said.
The expansion comes as the company seeks to increase the commercial reach of its AI hardware business.
Bloomberg Intelligence estimates that Google’s TPU sales could exceed $135 billion in 2028, compared with an estimated $12 billion this year.
AI disclosure: AI tools assisted with research and drafting of this piece. An Invezz editor reviewed the facts, sources and wording before publication. Any opinions or analysis are Invezz’s, not those of an automated system.
The post Alphabet stock rises as Google expands AI chip lineup to challenge Nvidia appeared first on Invezz
