GOOGLE CLOUD LAUNCHES NEW TPU CHIPS TO RIVAL NVIDIA
AI DESK■ 2 MIN READ
WED, APR 22, 2026
Google Cloud unveiled its latest generation of tensor processing units (TPUs), offering faster performance and lower costs than previous versions. The move intensifies competition with Nvidia while Google continues supporting Nvidia chips in its cloud services.
Google Cloud introduced two new AI chips designed to accelerate machine learning workloads and compete directly with Nvidia's dominant GPU offerings. The latest TPUs deliver improved performance metrics and reduced pricing, addressing customer demands for more cost-effective AI infrastructure.
The new chip lineup represents Google's ongoing effort to build proprietary hardware that differentiates its cloud platform. TPUs, first introduced in 2016, have become core components of Google's infrastructure for training and running large language models and other AI applications.
Despite launching its own chips, Google Cloud is maintaining support for Nvidia processors across its platform. This dual-hardware strategy allows customers to choose based on their specific workload requirements and preferences.
The release comes as demand for AI computing resources surges across industries. Companies are investing heavily in infrastructure to support generative AI applications, creating intense competition among cloud providers. Nvidia has maintained significant market share through its CUDA software ecosystem and first-mover advantage, but Google, Amazon, and Microsoft are all developing proprietary chips to reduce dependency and costs.
Google's new TPUs target customers running inference and training workloads on Google Cloud. The company has partnered with select enterprises to optimize applications for the new hardware, expanding adoption beyond internal Google use cases.
Analysts note that while Nvidia remains the market leader, Google's vertical integration of hardware and software gives it advantages in optimizing performance for specific AI tasks. The company can tailor chip architecture directly to TensorFlow and other frameworks it develops.
Google Cloud did not announce specific pricing or availability timelines for the new TPUs, though the company indicated broader rollout plans for the coming months. Early access partnerships are underway with enterprise customers across various sectors.
OpenAI has unveiled a dedicated physical keypad developed with keyboard maker Work Louder, designed to control AI agents. The device is available for order starting today.
Environmental and community groups are pushing for a pause on datacentre development in Australia until new regulations are finalized, following Prime Minister Anthony Albanese's announcement of an AI blueprint that includes energy rules for the sector.
Thinking Machines Lab has released Inkling, an open-source model with 975 billion parameters trained to process video and audio. The release positions the lab as a contender in the competitive AI landscape.
OpenAI's GPT-5.6 Sol Pro disproved a longstanding conjecture about the Benjamini-Hochberg statistical method in under two hours, where previous AI models and human researchers had failed for decades.