The Huawei Ascend 910 AI processor, built on the Da Vinci architecture with 7nm+ EUV technology, delivers 256 TFLOPS FP16 and 512 TOPS INT8 performance, superior energy efficiency, and a full-stack software ecosystem, making it ideal for large‑scale AI training, HPC, and cloud AI services.
The article provides a detailed technical analysis of Huawei's Ascend 910 AI processor, covering its Da Vinci architecture, hardware specifications, benchmark results, software ecosystem, application scenarios, and product roadmap, while also clarifying key terminology for readers.
The article explains what AI frameworks are, why they act as the operating system of artificial intelligence, showcases real‑world uses in transportation and finance, and provides an in‑depth analysis of Huawei's MindSpore framework, highlighting its development experience, hardware optimization, deployment flexibility, and enterprise‑grade security features.
This recap of the MindSpore public course reviews LLaMA2 fundamentals, compares its Transformer structure, details upgrades from LLaMA1, explains core components like RMSNorm, RoPE, KV‑Cache, Grouped Multi‑Query Attention and SwiGLU, outlines industry LLM optimization methods, and previews the upcoming lecture on the Pengcheng Brain 200B model.
This article reviews the LLaMA large‑language‑model series, covering its background, architectural innovations such as Add&Norm, SwiGLU, and RoPE, a known reversal‑curse bug, and provides step‑by‑step MindSpore Transformers code for model configuration, inference, and pipeline usage while previewing the upcoming LLaMA‑2 session.
This review of the recent MindSpore NLP public class walks through the fundamentals of large language model text generation, detailing deterministic decoding such as greedy and beam search, stochastic sampling techniques like temperature, top‑k and top‑p, and advanced methods including constrained beam, contrastive, and assisted search, with illustrative examples.
A university team built a skin tumor recognition system that identifies six malignant skin cancers with 91.6% accuracy, leveraging Huawei Cloud's ModelArts, MindSpore, and edge devices, while gaining valuable AI, cloud, and development experience and earning multiple competition awards.
During the Saturday session, we reviewed ChatGLM2’s upgrades—Multi‑Query Attention and FlashAttention—demonstrated deployment on Ascend + ModelArts + MindSpore, and introduced ChatGLM3’s revamped prompt design, native tool‑calling and code‑interpreter capabilities, while previewing the next lecture on text‑generation decoding.
The MindSpore Season 2 open class offers a comprehensive overview of GLM to ChatGLM architectures, positional‑embedding strategies, stable training optimizations, and step‑by‑step instructions for deploying large language models with Ascend, ModelArts, and MindSpore Transformers, while previewing upcoming multimodal remote‑sensing sessions.
During a livestream titled “Solving the ‘Development Difficulty’ of Large Models with MindSpore Auto Parallel”, Huawei’s MindSpore experts explained how the framework’s distributed training techniques—including data, model, and pipeline parallelism as well as memory‑saving strategies—enable efficient pre‑training of trillion‑parameter models across diverse AI domains.
The article introduces the MEGA-Protein tool, which integrates an AI‑driven MSA engine and the MindSpore framework to overcome AlphaFold 2’s limitations and achieve high‑accuracy protein structure prediction, and also announces the MSG Enterprise Tour and Hangzhou Developer Day event on April 27.
NetEase’s Fuxi team developed the multimodal ‘YuZhi’ model, a large‑scale image‑text dual‑tower system optimized with the EET inference framework, which powers personalized recommendations in NetEase News and Cloud Music, while a partnership with Huawei Ascend AI and MindSpore enables further model acceleration, compression, and the new ‘YuZhi‑Wukong’ model that improves video recommendation metrics by about 5%.
This article recounts how a former programmer became a deep‑learning evangelist, built Huawei’s MindSpore open‑source community, leveraged one‑minute videos, launched a hands‑on training camp and a lightweight tool, and fostered a thriving ecosystem that now serves over 190,000 developers.
