Google Unveils Next-Generation Tensor Processing Units for AI Inference
Google announced its next-generation Tensor Processing Units at Google Cloud Next, featuring dedicated inference chips designed to accelerate execution of trained AI models. The company is also developing custom memory processing units in collaboration with Marvell Technology to challenge NVIDIA's dominance.
New Chip Announcement
Google is set to unveil its next-generation Tensor Processing Units at Google Cloud Next this week, with dedicated inference chips designed to accelerate the execution of trained AI models
Strategic Partnerships
The company is also in talks with Marvell Technology to develop a memory processing unit and inference-optimized TPU, diversifying beyond Broadcom
Market Competition
Announced in reports on April 20, the push builds on recent deals supplying TPUs to Meta and Anthropic, targeting the exploding demand for efficient real-time AI responses
By challenging Nvidia's dominance in a market projected to grow rapidly, Google aims to lower costs and boost performance for its cloud customers, intensifying the chip wars and pressuring Nvidia while opening opportunities for custom silicon startups