--- 格式版本: 2 标题: "NVIDIA NVLink Fusion Expands With NVHBM Custom High-Bandwidth Memory" 原文链接: "https://blogs.nvidia.com/blog/nvlink-fusion-nvhbm-custom-high-bandwidth-memory/" 发布日期: "2026-08-26" 发布时间校准状态: "found" 发布时间需复核: "否" 发布时间来源: "rule:configured_publication_date_rule" 发布时间证据: "nvidia-blog-entry-date html:original: August 26, 2026" 发布时间校准原因: "信源发布日期识别规则直接确认发布时间" 发布时间校准置信度: "high" 发布时间候选数量: 2 发布时间严格候选数量: 2 发布时间原页读取状态: "source template page reused from URL open" 发布时间未找到原因: "" 发布时间校准时间: "2026-08-27T21:32:36+08:00" 发布时间仲裁状态: "skipped" 发布时间仲裁尝试次数: 0 发布时间仲裁耗时毫秒: 0 发现时间: "2026-08-27T21:31:55+08:00" 入库时间: "2026-08-27T13:32:36.426Z" 来源平台: "NVIDIA Blog 搜索" 搜索渠道: "source_template" 搜索词: "https://blogs.nvidia.com/?s=Scale-up" 匹配关键词: - "Scale-up" - "NVLink" - "bandwidth" - "XPU" - "HBM" - "performance" - "AI" 相关厂家: - "NVIDIA" - "AWS" 相关专家: [] 内容类型: "网页" 抓取工具: "Free Fetch + Defuddle" 清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取" 原始附件: [] AI优质: "是" AI打分: 85 AI分档: "高置信优质" AI质检状态: "通过" AI打分理由: "正文主线是NVLink Fusion机架级半定制AI基础设施及新型NVHBM，属于NVIDIA官方一手发布。历史证据已出现NVHBM基座裸片内置控制器及带宽、功耗、面积收益等核心信息，存在同事件重合；本文仍明确新增或强化了最高30%带宽提升、15%功耗降低、释放25%计算裸片面积，以及Annapurna Labs首批合作、Trainium4接入共同机架级架构等可核验事实。技术参数和集成路径完整，命中新产品/关键部件与具名客户合作通道；但合作仍偏未来规划，未披露量产、交付时间和规模，且发布日期缺失。" AI质检模型: "gpt-5.6-sol" AI质检时间: "2026-08-27T21:33:15+08:00" AI主题相关性: 20 AI来源权威性: 15 AI新颖性: 16 AI技术细节: 17 AI商业部署信号: 8 AI完整性: 9 AI评分提示词版本: "v17-精简生产版" AI评分提示词SHA256: "48fb9777f386026761b4873eaff30807694fb11e9b352d7c69bf2dfde750cc7d" AI评分知识库版本: "knowledge_base_v1-20260819+runtime.10" AI评分知识库SHA256: "72f38febd264ad2b85c2bb2b233a23b68d99610218a874843d0a5d12d9d97334" AI评分知识库检索词: "[\"NVIDIA\",\"Scale-up\",\"NVLink\",\"bandwidth\",\"https://blogs.nvidia.com/?s=Scale-up\",\"HBM\",\"HBM4\",\"rack-scale\",\"RAS\",\"GPU\",\"XPU\",\"NVHBM\"]" AI评分知识库命中: "[{\"id\":\"runtime-676f90ee5f88ebab3b504077\",\"title\":\"1 NVIDIA NVLink Fusion Brings NVHBM to Next-Generation AI Infrastructure | August 2026 NVIDIA's NVLink Fusion technology enables hyperscalers and AI natives to deploy custom XPUs and CPUs into the NVIDIA AI infrastructure platform, while NVHBM, a custom HBM base-die technology, increases memory bandwidth, reduces power consumption, and provides more package and silicon area for compute, as discussed by speakers Jesse Clayton, Varun Nanda Kumar, and Farshad Ghodsian, allowing for more efficient a\",\"sourceType\":\"ai_excellent_article\",\"time\":\"\",\"matchedTerms\":[\"NVIDIA\",\"Scale-up\",\"NVLink\",\"bandwidth\",\"HBM\",\"HBM4\",\"rack-scale\",\"RAS\",\"GPU\",\"XPU\",\"NVHBM\"],\"rank\":-33.69808948331113},{\"id\":\"july-correct-0095\",\"title\":\"Inside NVIDIA Rubin GPU Architecture: Powering the Era of Agentic AI\",\"sourceType\":\"labeled_article\",\"time\":\"2026-07\",\"matchedTerms\":[\"NVIDIA\",\"Scale-up\",\"NVLink\",\"bandwidth\",\"HBM\",\"HBM4\",\"rack-scale\",\"RAS\",\"GPU\"],\"rank\":-18.972174423078904},{\"id\":\"july-correct-0089\",\"title\":\"Setting a World Record for MoE Pre-Training on NVIDIA GB300 NVL72\",\"sourceType\":\"labeled_article\",\"time\":\"2026-07\",\"matchedTerms\":[\"NVIDIA\",\"Scale-up\",\"NVLink\",\"bandwidth\",\"HBM\",\"rack-scale\",\"RAS\",\"GPU\"],\"rank\":-14.619632751404026},{\"id\":\"july-correct-0079\",\"title\":\"AAI 2026: AMD Launches AMD Helios Rackscale Solution for Frontier AI\",\"sourceType\":\"labeled_article\",\"time\":\"2026-07\",\"matchedTerms\":[\"NVIDIA\",\"Scale-up\",\"bandwidth\",\"HBM\",\"HBM4\",\"rack-scale\",\"RAS\",\"GPU\"],\"rank\":-14.425862161298465},{\"id\":\"runtime-4abbccc42d96af674efc7768\",\"title\":\"OpenAI Jalapeño: Better Than Nvidia Blackwell\",\"sourceType\":\"ai_excellent_article\",\"time\":\"2026-08-25\",\"matchedTerms\":[\"NVIDIA\",\"Scale-up\",\"bandwidth\",\"HBM\",\"HBM4\",\"rack-scale\",\"RAS\",\"GPU\",\"XPU\"],\"rank\":-13.465777058115904}]" AI摘要: "NVIDIA宣布扩展NVLink Fusion，推出NVHBM定制高带宽内存，将NVIDIA定制内存控制器集成到HBM堆栈中，较标准HBM4E提升30%内存带宽并降低15%功耗。" AI摘要模型: "ali-deepseek-v4-flash" AI摘要时间: "2026-08-27T18:59:48.192Z" 采集批次: "2026年8月27日21点31分48秒" 采集批次ID: "20260827-213148-742" 去重键: "https://blogs.nvidia.com/blog/nvlink-fusion-nvhbm-custom-high-bandwidth-memory" --- The next wave of AI is placing new demands on infrastructure. As AI agents and trillion-parameter workloads become mainstream, the performance of AI infrastructure depends not only on compute, but on how compute, memory, storage, networking and software are designed together as a unified system. To help hyperscalers and AI innovators build the next generation of semi-custom AI infrastructure, NVIDIA today expanded [NVIDIA NVLink Fusion](https://www.nvidia.com/en-us/data-center/nvlink-fusion/) with [NVIDIA NVHBM](http://developer.nvidia.com/blog/nvidia-nvlink-fusion-brings-nvhbm-to-next-generation-ai-infrastructure), a next-generation high-bandwidth memory technology that brings higher memory performance and efficiency to XPUs. It will be validated and offered by leading memory partners, extending this advanced memory capability to NVLink Fusion customers. Traditional HBM architectures place the memory controller on the XPU die, consuming valuable silicon area that could otherwise be dedicated to compute. NVHBM, built on the same technology that NVIDIA will use for future GPUs, integrates NVIDIA’s custom memory controller into the HBM base die. By integrating the memory controller into the 3D HBM stack instead of the XPU, NVHBM delivers up to 30% greater memory bandwidth and 15% lower HBM power consumption, and frees up to 25% more area on XPU compute die compared with standard HBM4E. NVIDIA is establishing a standard NVHBM implementation, available from multiple memory providers. This reduces the engineering effort required to integrate and qualify memory across multiple suppliers — giving NVLink Fusion customers a faster path for bringing custom AI chips to market. [Amazon’s Annapurna Labs](https://nvidianews.nvidia.com/news/aws-and-nvidia-to-deliver-2-million-additional-gpus-and-next-generation-infrastructure-for-agentic-and-physical-ai) will be the first to work on NVHBM as part of its broader collaboration with NVIDIA around NVLink Fusion. ## AWS and NVIDIA Continue NVLink Fusion Collaboration Amazon’s Annapurna Labs will work with NVIDIA on NVHBM technology and the NVLink scale-up architecture to enhance performance and efficiency for AI workloads. This builds on AWS’s [previously announced](https://blogs.nvidia.com/blog/aws-partnership-expansion-reinvent/) support for NVLink Fusion. Annapurna Labs will support NVLink Fusion with its next-generation Trainium chips starting with Trainium4, which would allow Amazon chips and NVIDIA GPUs to work together with common rack-scale architecture. “NVHBM represents a new architectural approach to advancing high-bandwidth memory performance and efficiency,” said Nafea Bshara, vice president of Annapurna Labs at Amazon. “We look forward to this technology collaboration to benefit future AWS infrastructure designs.” ## Vertically Integrated and Horizontally Open NVLink Fusion enables partners to connect custom XPUs and CPUs to NVIDIA’s rack-scale platform. Partners can access NVIDIA NVLink chiplets, NVLink-C2C, NVLink Switches and NVIDIA MGX systems and racks, as well as a broad ecosystem of CPU partners, ASIC designers, system manufacturers and technology providers. Offered with each generation of NVIDIA’s rack-scale system architecture, NVLink Fusion allows hyperscalers and AI-native companies to focus engineering resources on XPU innovation while using a proven technology stack for scale-up and scale-out networking, rack-scale systems and software — creating a faster, lower-risk path to deploying semi-custom AI infrastructure. *Learn more about* [*NVLink*](https://www.nvidia.com/en-us/data-center/nvlink/) *and* [*NVLink Fusion*](https://www.nvidia.com/en-us/data-center/nvlink-fusion/)*.*