The new TPU 8tx and TPU 8i Max expand Google's eighth-generation chip lineup that originally debuted in April. Developed with AMD, the TPU 8tx features a custom inferencing network that enables simultaneous reasoning alongside training while maintaining hardware-level protections for sensitive model weights. Meanwhile, the TPU 8i Max combines two TPU 8i chips into a single package to double processing power, memory, and bandwidth for intensive query workloads. This roll-out comes as market leader Nvidia faces fresh scrutiny following its chief executive's admission of an affair with an alleged Chinese spy.
Key points
Google is expanding its proprietary silicon to decrease its reliance on third-party semic…: Google is expanding its proprietary silicon to decrease its reliance on third-party semiconductor manufacturers.
Major artificial intelligence laboratories including Microsoft, Amazon, OpenAI, and Anthr…: Major artificial intelligence laboratories including Microsoft, Amazon, OpenAI, and Anthropic are actively building rival proprietary chips to curb Nvidia's dominance.
The new hardware specifically addresses advanced cloud infrastructure demands for confide…: The new hardware specifically addresses advanced cloud infrastructure demands for confidential inference and large-scale, memory-intensive data processing.
Sources checked by our analysis pipeline: 1
Google is notching up its chipmaking game as the company seeks to reduce its dependence on chips made by other companies for its cloud computing business. Google on Monday unveiled two new variants of its tensor processing units (TPUs), in-house developed chips meant for AI workloads, that broaden its chip portfolio to additional use cases: confidential inference and more memory capacity for large workloads.
The new units, TPU 8tx and TPU 8i Max, are part of the company’s eighth generation of chips introduced in April this year, and build on what those chips offered, which was more inference capacity and support for large training requirements.
TPU 8tx, designed with AMD, is based on the TPU 8t chip, and is geared specifically for confidentially training and running AI models, Google said in a blog post. The chip “builds security directly into the hardware layer, keeping sensitive data and model weights protected and isolated even while in active use,” the company wrote.
The chip also has a custom inferencing network that lets it run reasoning alongside training, so it can be used…
Wall Street projections suggest a steep financial trajectory for the artificial intelligence firm over the coming decade. The long-term models anticipate massive top-line expansion alongside a delayed timeline for bottom-line profitability. Achieving these benchmarks would position the business among the highest-earning technology enterprises globally.
The launch targets advanced technical workflows like multi-step programming and scientific discovery following an exclusive early preview for cybersecurity specialists. By pairing the powerful architecture with multi-agent coordination on the Antigravity platform, the tech giant aims to capture high-end enterprise demand. This release intensifies industry competition by heavily undercutting rival developer pricing while delivering performance near the top of public benchmarks.
Investors have pushed the firm's private market worth to new heights amid sustained enthusiasm for specialized machine learning infrastructure. The enterprise specializes in workforce training solutions specifically designed for modern artificial intelligence architectures. Achieving this multi-billion-dollar scale highlights the immense value placed on companies preparing data and human capital for next-generation models.