Build intelligent computing infrastructure

Specializing in AI server R&D and manufacturing, delivering high-performance and highly reliable computing solutions for enterprises and government agencies.

Built on domestic CPU platforms including Kunpeng, Hygon, Phytium, and LoongArch; certified under ISO 9001 quality management system; supports cutting-edge technologies such as PCIe 5.0 and DDR5.

Trusted Partner

Trusted by leading enterprises and government agencies as their AI infrastructure partner.

"We employ a full lifecycle quality management approach. From component screening to final unit testing, every stage adheres to rigorous quality control standards, ensuring reliable performance even in complex environments." — Zhang, Technical Director

5000+

Cumulative AI Server Deliveries

Data source: Internal delivery system, as of Q4 2024

200+

Serve enterprise clients

Covers government, finance, education, and other sectors, including numerous central state-owned enterprises.

99.99%

Runtime Stability

Complies with GB/T 2887-2011 standard based on MTBF testing.

Core Product Line

Delivers comprehensive Xinchuang computing solutions across all scenarios, built on domestic CPU platforms including Kunpeng, Hygon, Phytium, and Loongson.

Hygon Server

Hygon Server

High-performance general-purpose computing and AI training on Hygon CPU platforms

Supports PCIe 4.0/5.0, DDR4/DDR5; up to 2TB memory

Technical Approach:Features multi-channel memory architecture and intelligent power management, ensuring system stability through rigorous compatibility testing.

View details
Kunpeng Server

Kunpeng Server

Built on the Kunpeng 920 platform with deep integration into openEuler and HarmonyOS ecosystems.

ARMv8 architecture, 64-core processor with NUMA optimization support

Technical Approach:Optimize NUMA topology for ARM architectures and enhance resource utilization through hardware-level virtualization support.

View details
Phytium Server

Phytium Server

First choice for high-security government and military scenarios on Phytium platforms.

Meets Class 3 MLPS certification; supports SM cryptographic algorithms; secure boot enabled

Technical Approach:Leverages Trusted Computing 3.0 technology to enable secure boot and runtime protection at the hardware level, meeting MLPS Level 3 requirements.

View details
Loongson Server

Loongson Server

Built on LoongArch, a fully independent and controllable instruction set.

LoongArch Instruction Set, Independent Ecosystem, Supports Domestic OS

Technical Approach:Designed around a proprietary instruction set architecture to achieve full technical independence at the foundational level, building a complete ecosystem alongside domestic operating systems.

View details
X86 Server

X86 Server

General-purpose computing power for enterprise core business on Intel/AMD platforms

Intel Xeon/AMD EPYC, supports PCIe 5.0, CXL 2.0

Technical Approach:Leverage the latest CXL 2.0 interconnect technology for memory pooling and resource sharing to boost data center efficiency.

View details
GPU Server

GPU Server

High-performance GPU clusters covering all AI training and inference scenarios

Supports NVIDIA and domestic GPUs, NVLink interconnect, and liquid cooling

Technical Approach:Leverages NVLink high-speed interconnects and liquid cooling technology to enable large-scale parallel computing via GPU cluster optimization.

View details
Frequently Asked Questions

Frequently Asked Questions: Products & Customization

The following answers address the most frequently asked questions regarding actual selection and customization. For project-specific assessments, please contact our technical team.

What types of AI servers does Baisun Huagong offer?+

Built on CPU platforms from Hygon, Kunpeng, Phytium, Loongson, Intel, and AMD, our products cover AI training servers, AI inference servers, general-purpose rack servers, edge computing servers, and workstations. They support NVIDIA and domestic GPU acceleration cards.

Does the product support domestic localization and Xinchuang requirements?+

Our flagship products are built on domestic CPU platforms from Hygon, Kunpeng, Phytium, and Loongson. They support adaptation to domestic operating systems and software ecosystems and can provide certification and testing materials as required by the project.

Do you offer OEM/ODM customization services?+

End-to-end customization from motherboard and chassis to complete units, including configuration, branding, design & structure, functional features, and certification compatibility. Tailored product definitions based on your business scenarios.

How is the delivery timeline for a custom project determined?+

Milestones are defined across four phases: design, prototyping, validation, and mass production. Actual timelines depend on customization depth and batch size, as specified in the written proposal and contract.

Contact Now

Leave your contact information, and our professional consultant will reach out to you within 24 hours.

Contact Us

  • Sales Hotline

    189-2600-9050
  • Customer Support Email

    mkt@100trust.com.cn

  • Company Address

    Building 1, Floor 14, Phase 2, Shenzhen Software Park, Tech Middle 2nd Road, Yuehai Street, Nanshan District, Shenzhen