---
格式版本: 2
标题: "Build Semi-Custom AI Infrastructure |  NVIDIA NVLink Fusion"
原文链接: "https://www.nvidia.com/en-us/data-center/nvlink-fusion"
发布日期: "未标注"
发现时间: "2026-05-25T00:00:00+08:00"
入库时间: "2026-05-28T02:39:39+08:00"
来源平台: "历史资料迁移"
搜索渠道: "legacy_migration"
搜索词: "site:nvidia.com/en-us/data-center NVL72 OR \"AI factory\" OR \"rack\" after:2026-05-18"
匹配关键词:
  - "AI Rack"
  - "厂商官网"
  - "高质量信源"
  - "新资料"
  - "site:nvidia.com/en-us/data-center NVL72 OR \"AI factory\" OR \"rack\" after:2026-05-18"
  - "待定数据"
  - "厂家"
  - "Nvidia"
  - "Build"
  - "Semi"
  - "Custom"
  - "AI"
  - "Infrastructure"
  - "NVIDIA"
  - "NVLink"
  - "Fusion"
  - "25.md"
相关厂家:
  - "NVIDIA"
  - "AWS"
  - "Google"
相关专家:
  []
内容类型: "网页"
抓取工具: "Firecrawl 原始抓取 + qwen3.6-flash 原文清洗"
清洗工具: "历史资料原文保留 + LLM 正文裁剪"
关联判断: "待人工确认"
关联置信度: 0
关联理由: "历史资料迁移，未重新调用模型判定"
AI优质: "是"
AI打分: 93
AI分档: "高置信优质"
AI质检状态: "通过"
AI打分理由: "官方发布NVLink Fusion机柜级AI平台，明确涵盖NVL72架构、NVLink 6互连规格（72卡/3.6TB/s）、CPO及OCP MGX标准，并披露AWS Trainium4部署进展与完整生态链，技术细节与商业信号极强。"
AI质检模型: "qwen3.6-plus"
AI质检时间: "2026-06-03T00:19:45+08:00"
AI主题相关性: 20
AI来源权威性: 15
AI新颖性: 18
AI技术细节: 17
AI商业部署信号: 14
AI完整性: 9
采集批次: "2026年5月28日2点39分33秒"
采集批次ID: "legacy-migration-20260528023933"
去重键: "https://www.nvidia.com/en-us/data-center/nvlink-fusion"
---

# Build Semi-Custom AI Infrastructure | NVIDIA NVLink Fusion

[Read Press Release](https://nvidianews.nvidia.com/news/nvidia-nvlink-fusion-semi-custom-ai-infrastructure-partner-ecosystem)

### Overview

## World Class Scale-Up Performance
NVIDIA NVLink™ Fusion is a rack-scale AI infrastructure platform that enables hyperscalers and custom ASIC designers to integrate custom CPUs and XPUs with the world-leading NVLink scale-up interconnect and OCP MGX rack-scale server architecture. Leverage NVIDIA’s battle-tested AI technology stack and proven rack-scale design and ecosystem to reach the market faster, reduce development costs and deployment risks, and achieve better performance for a higher return on investment. The NVLink Fusion modular technology portfolio includes NVIDIA GPUs, NVIDIA Vera™ CPUs, NVLink scale-up networking, co-packaged optics (CPO) switches, NVIDIA ConnectX® SuperNICs, BlueField® DPUs, and Mission Control software. The comprehensive ecosystem includes ASIC designers, CPU and IP providers, OEMs/ODMs, and component suppliers, all designed to enable rapid deployment of custom AI silicon.

### AWS Integrates AI Infrastructure with NVIDIA NVLink Fusion for Trainium4 Deployment
Learn how AWS is using NVLink Fusion to accelerate Trainium4 deployment. [Read Blog](https://developer.nvidia.com/blog/aws-integrates-ai-infrastructure-with-nvidia-nvlink-fusion-for-trainium4-deployment/)

### Integrating Semi-Custom Compute into Rack-Scale Architecture with NVIDIA NVLink Fusion
Learn how NVIDIA NVLink Fusion allows hyperscalers to build semi-custom AI infrastructure, integrating their ASICs or CPUs with NVIDIA GPUs, while standardizing on a single scalable hardware infrastructure. [Read Blog](https://developer.nvidia.com/blog/integrating-custom-compute-into-rack-scale-architecture-with-nvidia-nvlink-fusion)

## Benefits

### World-Class Scale-Up Performance
Unlocking the full potential of AI factories requires swift, seamless communication among all accelerators. NVIDIA NVLink 6 can connect 72 XPUs all-to-all at 3.6 TB/s per XPU to boost AI performance and return on investment.

### Lower Development Costs
The established NVLink Fusion supplier ecosystem provides all of the components required for full rack-scale deployment based on the OCP MGX architecture, from rack and chassis to power delivery and cooling systems, eliminating the development costs and deployment risks associated with a new rack design.

### Accelerated Time to Market
By leveraging NVIDIA’s battle-tested technology stack and ecosystem of ASIC designers, CPU and IP providers, and OEM/ODMs, hyperscalers can achieve faster time to market and faster time to revenue.

### Single Unified Architecture
As hyperscalers already deploy full NVIDIA rack solutions, NVLink Fusion allows heterogeneous silicon offerings while standardizing around a common rack design, accelerating AI factory deployment, and simplifying management.

## Platform

### NVIDIA NVLink
NVIDIA NVLink 6 and NVLink Switch Chip enable 260 TB/s of bandwidth in a single 72-accelerator NVLink domain (NVL72) and deliver 4x bandwidth efficiency with NVIDIA Scalable Hierarchical Aggregation and Reduction Protocol (SHARP)™ FP8 support. [Learn More About NVIDIA NVLink and NVLink Switch](https://www.nvidia.com/en-us/data-center/nvlink/)

### NVIDIA NVLink-C2C
NVIDIA NVLink-C2C extends the industry-leading NVLink technology to a chip-to-chip interconnect. This enables the creation of a new class of integrated products with NVIDIA partners, built via chiplets, allowing NVIDIA GPUs or CPUs to have a high-bandwidth coherent connection with custom silicon. [Learn More About NVIDIA NVLink-C2C](https://www.nvidia.com/en-us/data-center/nvlink-c2c/)

## Adopters

## NVLink Fusion Ecosystem
2. CPU Partners
3. Custom Silicon Partners
4. Technology Partners
![arm logo](./assets/09-www.nvidia.com.jpg)
![Intel logo](./assets/10-www.nvidia.com.jpg)
![Fujitsu logo](./assets/11-www.nvidia.com.jpg)
![Qualcomm logo](./assets/12-www.nvidia.com.jpg)
![SiFive logo](./assets/13-www.nvidia.com.jpg)
![alchip logo](./assets/14-www.nvidia.com.jpg)
![Astera Labs logo](./assets/15-www.nvidia.com.jpg)
![GUC logo](./assets/16-www.nvidia.com.jpg)
![Marvell logo](./assets/17-www.nvidia.com.jpg)
![Mediatek logo](./assets/18-www.nvidia.com.jpg)
![Samsung logo](./assets/19-www.nvidia.com.jpg)
![Cadence logo](./assets/20-www.nvidia.com.jpg)
![Synopsys logo](https://www.nvidia.com/content/nvidiaGDC/us/en_US/data-center/nvlink-fusion/_jcr_content/root/responsivegrid/nv_container_1153353446/nv_tabs/item_1760101795597/nv_container_copy_co/nv_image_copy_184311_1820289429.coreimg.svg/1764743824742/synopsys-logo.svg)

## Scaling AI Inference Performance with NVLink Fusion
Learn how NVIDIA NVLink Fusion addresses the growing demands of complex AI models. [Read Blog](https://developer.nvidia.com/blog/scaling-ai-inference-performance-and-flexibility-with-nvidia-nvlink-and-nvlink-fusion/)

## 原文链接
https://www.nvidia.com/en-us/data-center/nvlink-fusion