SOINTECH / INSIGHTS

SOIN Technology und ESWIN Computing: RISC-V+AI-All-in-One-Systeme als digitale KI-Infrastruktur für Unternehmen

SOIN Technology has officially become a reseller of ESWIN Computing. The two parties will cooperate on enterprise AI computing all-in-one systems and RISC-V+AI computing products, delivering localized intelligent computing solutions that help enterprises build secure and controllable AI digital infrastructure.

2025 RISC-V Industry Development Conference and RDSA International Forum

About ESWIN Computing

ESWIN Computing is a chip product provider based on the RISC-V architecture. Following its RISC-V+AI technology roadmap, its portfolio spans human-machine interaction, multimedia, connectivity and computing, serving consumer electronics, automotive electronics, robotics and industrial control. On October 9, 2026, ESWIN Computing was officially listed on the Main Board of the Hong Kong Stock Exchange (stock code 01256.HK). As of March 31, 2026, the company held more than 620 IP modules and over 20 series-grade RISC-V processor cores, with more than 1,740 patent applications, over 150 hardware-software co-design products and 220 customers worldwide. It is a Board member of RISC-V International and initiated the RDI (RISC-V Digital Infrastructure) Alliance.

ESWIN Computing listing ceremony at The Stock Exchange of Hong Kong

RISC-V+AI: Intelligent Computing for the Enterprise

In its computing product line, the ESWIN EIC77 series of intelligent computing SoCs has drawn wide industry attention. The EIC7702X integrates an 8-core 64-bit out-of-order RISC-V CPU, self-developed NPUs and a high-performance 3D GPU, delivering up to 40 TOPS of INT8 compute with full-stack floating-point support. It supports on-device inference of mainstream large models such as DeepSeek, offers energy-efficient audio/video codec capabilities, and provides rich peripherals including USB, PCIe, HDMI and Ethernet.

ESWIN Computing at Auto China 2026, showcasing automotive-grade chip solutions based on its self-developed RISC-V cores

Based on the EIC77 series, ESWIN Computing offers chips, development boards, SOM/OSM modules, SBC single-board computers, AI accelerator/video transcoding cards, AI Boxes, AI PCs and edge intelligent stations, covering cloud, edge and endpoint deployment scenarios and supporting mainstream operating systems such as Ubuntu, Debian and openKylin.

ESWIN Computing booth showcasing RISC-V+AI product forms

AI Digital Infrastructure for Enterprises

As large models move from the cloud to on-premises deployment, enterprises demand intelligent computing that is secure, cost-controllable and reliably available. AI computing all-in-one systems built on ESWIN Computing's RISC-V+AI products can run large-model inference, knowledge bases and agent applications locally within the enterprise — data stays in-domain and the systems work out of the box, providing a stable and manageable AI digital infrastructure.

As an ESWIN Computing reseller, SOIN Technology will provide computing sizing and solution design, all-in-one system delivery and deployment, large-model and agent application adaptation, and ongoing operations support, helping customers in manufacturing, energy, government and education accelerate their intelligent upgrades.

ESWIN Computing automotive showcase covering image processing, intelligent transmission, intelligent control and intelligent interaction
ESWIN Computing (01256.HK) Main Board listing on HKEX, with ecosystem partners on site

Building an Open Ecosystem Together

SOIN Technology stated that this partnership is a key step in its enterprise AI infrastructure strategy. Going forward, SOIN Technology will work with partners such as Baidu AI Cloud and ESWIN Computing to connect the full stack from intelligent computing infrastructure to agent applications, offering enterprise customers a complete path from computing power to real-world AI applications.

Dieser Inhalt stammt aus öffentlichen Internetmedien und gibt nicht die Auffassung der Beijing SOIN Technology Co., Ltd. wieder.