Tag
OpenAI has published lab results for its custom ASIC Jalapeño, showing it outperforms NVIDIA's GB200 and GB300 chips and rivals Vera Rubin in inference efficiency.
Google is reportedly collaborating with AMD to design a next-generation TPU that could integrate CPU cores on-package for reinforcement learning workloads, potentially advancing AI hardware architecture.
Zhipu AI is exploring a custom ASIC after DeepSeek, as GLM-5.2 usage jumped 27x in one week, highlighting the shift towards specialized hardware for inference at scale in China's AI industry.