Google Cloud launches two new AI chips to compete with Nvidia

Google Cloud announced its eighth-generation AI chips (TPU 8t for training, TPU 8i for inference) delivering 3x faster training, 80% better cost-per-performance, and ability to cluster 1M+ units together, positioning it as a significant alternative to Nvidia while reducing energy and compute costs for enterprises. However, Google is augmenting rather than replacing Nvidia infrastructure and continues partnering with Nvidia, suggesting a hybrid multi-chip strategy will dominate cloud AI deployments in the near term. CIOs should recognize this signals increased AI chip optionality and potential cost savings for cloud workloads, but Nvidia dominance in the AI infrastructure market is unlikely to erode quickly.