From GPU to NPU: Solving the AI Computing Dilemma in Mobile Chips

From GPU to NPU: Solving the AI Computing Dilemma in Mobile Chips

This article is compiled by the Semiconductor Industry Review (ID: ICVIEW) from semiengineering. Edge AI, GenAI, and next-generation communications are adding more workloads to smartphones, which are already under pressure to deliver high performance with low power consumption. Leading smartphone vendors are striving to keep up with the growing computational and power demands driven by … Read more

The Feast of Edge AI is Taking Shape

The Feast of Edge AI is Taking Shape

Looking back at the year 2024, artificial intelligence (AI) technology has rapidly developed in fields such as mobile devices, personal computers, automotive intelligent driving, digital home appliances, smart healthcare, and industrial applications, becoming the core driving force for innovation and development in multiple areas, providing strong support for the development of almost all sectors. As … Read more

Understanding ASIC, CPU, FPGA, CGRA, and GPU Architectures

Understanding ASIC, CPU, FPGA, CGRA, and GPU Architectures

ASIC Application-Specific Integrated Circuit. An ASIC is designed for a specific application, optimized for a particular task or function, but its wiring and connections are fixed and cannot be changed. For example, Bitcoin mining chips and mobile image processing units. CPU Handles general tasks. A CPU operates using SIMD (Single Instruction, Multiple Data), which is … Read more

Deployment of vLLM Enterprise Large Model Inference Framework (Linux)

Deployment of vLLM Enterprise Large Model Inference Framework (Linux)

Introduction Compared to traditional LLM inference frameworks (such as HuggingFace Transformers, TensorRT-LLM, etc.), vLLM demonstrates significant advantages in performance, memory management, and concurrency capabilities, specifically reflected in the following five core dimensions:1. Revolutionary Improvement in Memory Utilization By utilizing Paged Attention technology (inspired by the memory paging mechanism of operating systems), the KV Cache (Key-Value … Read more

Defeating AMD and Nvidia: Apple’s Path to Dominance in the Mobile GPU Market

Defeating AMD and Nvidia: Apple's Path to Dominance in the Mobile GPU Market

From 2006 to 2013, AMD and Nvidia completely mismanaged their market competition in mobile platforms, losing their status as major global GPU suppliers while Apple gradually replaced them to become the most powerful and mainstream producer of GPU processors. This article reviews this history to analyze why Apple can still replicate its success, hoping to … Read more

Virtualization Tutorial (10): Guide to Using the NVIDIA Display Mode Selector Tool

Virtualization Tutorial (10): Guide to Using the NVIDIA Display Mode Selector Tool

NVIDIA GPUs provide powerful graphics processing and computing capabilities for enterprises and developers. However, to fully leverage the potential of the GPU, proper display port configuration is crucial.NVIDIA Display Mode Selector Tool was created for this purpose, helping you easily enable/disable physical display ports, optimize GPU graphics processing, computing performance, and virtual GPU deployment, ensuring … Read more

Comparison of Rockchip RK3399 and Intel Celeron J1900

Comparison of Rockchip RK3399 and Intel Celeron J1900

Processor Basic Information Processor Rockchip RK3399 Intel Celeron J1900 Main Market Single Board Computer Mini Desktop Architecture ARMv8 – A (64-bit) x86 – 64 (64-bit) Release Date Q1 2016 Q4 2013 Process Technology 28nm HKMG 22nm Number of Cores 6 4 Number of Threads 6 4 Base Frequency 1.5GHz 2.0GHz Turbo Frequency 2.0GHz 2.42GHz Cache … Read more

Insights: GPU or ASIC, Which Will Drive the Large Language Model’s Scalable Development?

Insights: GPU or ASIC, Which Will Drive the Large Language Model's Scalable Development?

Large Language Models (LLMs) are just getting started. The CEOs of OpenAI, Anthropic, and xAI share remarkably similar visions—exponential growth in artificial intelligence will transform humanity, with impacts far exceeding most people’s expectations. This is not just speculation. Today, the market and value of artificial intelligence have become a reality: Human developers using GitHub CoPilot … Read more

MediaTek: Dimensity 1200 Project Started in 2019, Mobile Ray Tracing is Here

MediaTek: Dimensity 1200 Project Started in 2019, Mobile Ray Tracing is Here

On January 20, MediaTek held an online launch event for the Dimensity series, officially unveiling the new generation flagship 5G chip – Dimensity 1200.Compared to the previous generation Dimensity 1000 series, the new Dimensity 1200 has made significant improvements in CPU/GPU performance, 5G, AI, photography, video, and gaming.The Dimensity 1200 chip is manufactured using TSMC’s … Read more

In-Depth Analysis of Samsung Exynos 9820: The Android Super Chip?

In-Depth Analysis of Samsung Exynos 9820: The Android Super Chip?

On November 14, Beijing time, Samsung officially launched its next-generation flagship mobile platform, the Exynos 9820. As the name suggests, it is an iterative version of the previous Exynos 9810, and it is expected that some versions of next year’s Galaxy S10 and Note10 will be equipped with this chip. How powerful is the Samsung … Read more