AMD's Q3 earnings revealed a 97% surge in data center revenue to $8.55B, driven by Instinct GPUs and EPYC processors. The company raised its 2027 revenue growth forecast to over 80% and guided for the Helios rack system to scale production in Q4, with Meta already testing initial units.
AI Accelerator
AI 加速器
Specialized hardware optimized for artificial intelligence workloads
Curated Intelligence
IDC data reveals Nvidia's China AI accelerator shipments fell 44% year-over-year in the first half of the year, dropping its market share to 34%. Conversely, Huawei captured a 28% share with 572,000 units shipped, while domestic players like Alibaba's T-Head and Cambricon also saw significant growth, indicating a structural shift in China's compute landscape.
OpenAI and Broadcom announced a strategic collaboration to deploy 10 gigawatts of custom AI accelerators designed by OpenAI and developed by Broadcom. The partnership aims to embed frontier model insights directly into hardware to unlock new capabilities, with rack deployments starting in H2 2026 and completing by end of 2029.
Hugging Face announced the native integration of Intel Gaudi hardware support into the main codebase of Text Generation Inference (TGI). This update provides the open-source community with production-ready deployment options using specialized AI accelerators to optimize LLM inference performance and flexibility.
Intel demonstrates running the ProtST protein language model on Gaudi 2 chips. The case primarily validates hardware compatibility without providing specific performance uplift data or benchmarks.
The source text contains only a title regarding the text-generation pipeline on Intel Gaudi 2, lacking any technical details, performance metrics, or implementation methods. Due to the complete absence of core facts, its value and usability cannot be assessed.
The article title indicates acceleration of the BridgeTower vision-language model on Habana Gaudi2 chips. Due to missing body content, specific performance metrics, implementation methods, or technical details cannot be evaluated.
Habana Labs and Hugging Face announced a partnership to integrate Habana Gaudi AI processors into the Transformers library for accelerated model training. The initiative aims to reduce compute costs and complexity for large-scale Transformer models by offering an alternative hardware solution to NVIDIA GPUs.