Scale Computing Hypercore

Huawei 384 Optical Module Computing Power

Huawei's CloudMatrix 384 Supernode, powered by 384 Ascend 910C chips, rivals Nvidia's GB200 NVL72 with 300 petaflops of AI compute power. Explore its impact on global AI and China's tech self-sufficiency. 2% failures stem from optics & how QSFPTEK cuts costs by 69. On May 14, 2025, the "2025 Chip and Optical Forum" hosted by HiSilicon and organized by. In the AI era, Huawei provides a full range of GE to 800GE optical modules, featuring three major capabilities: Spanning (ultra-long transmission), Stable (ultra-high reliability), and Secure (ultra-solid security). Huawei Technologies has introduced the CloudMatrix 384 Supernode, a groundbreaking AI. Huawei recently started delivering its new CloudMatrix 384 AI clusters to Chinese customers – and is making no secret of its goal: technological independence from Western suppliers, particularly NVIDIA.

[PDF Version]

Huawei Intelligent Computing Center Short-Range Optical Module

At MWC 2025, Huawei introduced the StarryLink Series optical modules – a breakthrough solution designed to keep data flowing effortlessly into the future. As AI technology advances, the need to collate and process scattered parameters has become increasingly critical. In the AI era, Huawei provides a full range of GE to 800GE optical modules, featuring three major capabilities: Spanning (ultra-long transmission), Stable (ultra-high reliability), and Secure (ultra-solid security). With AI models processing unprecedented volumes of parameters, the need for highly reliable. With the surge in AI development, AI training clusters have evolved to a scale of 10,000+ GPUs, resulting in a significant increase in the number of optical modules required.

[PDF Version]