---
格式版本: 2
标题: "Designed for AI Reasoning Performance & Efficiency | NVIDIA GB300 NVL72"
原文链接: "https://www.nvidia.com/en-gb/data-center/gb300-nvl72/"
发布日期: "2026-06-12"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:scrape:strict_markdown_body"
发布时间证据: "[ June 12, 2026 NVIDIA Achieves Leading Agentic Coding Performance on First Agentic AI Benchmark ]("
发布时间校准原因: "规则确认唯一严格发布时间，来源 scrape:strict_markdown_body"
发布时间校准置信度: "high"
发布时间候选数量: 5
发布时间严格候选数量: 1
发布时间原页读取状态: "原页面已读取"
发布时间未找到原因: ""
发布时间校准时间: "2026-07-24T21:07:58+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: "2026-06-13T08:30:47+08:00"
入库时间: "2026-07-24T13:07:58.900Z"
来源平台: "Exa"
搜索渠道: "exa_search"
搜索词: "NVL72"
匹配关键词:
  - "NVL72"
相关厂家:
  - "NVIDIA"
相关专家:
  []
内容类型: "网页"
抓取工具: "XCrawl Scrape"
清洗工具: "XCrawl Markdown + LLM 正文裁剪（重试成功）"
原始附件:
  []
AI优质: "是"
AI打分: 95
AI分档: "高置信优质"
AI质检状态: "通过"
AI打分理由: "NVIDIA官方页面，直接详细描述GB300 NVL72机柜级AI系统架构，包含GPU/CPU/NVLink、HBM、液冷、性能参数等关键技术细节，且为最新发布产品，高置信优质。"
AI质检模型: "deepseek-v4-flash"
AI质检时间: "2026-07-24T21:32:05+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: 20
AI商业部署信号: 12
AI完整性: 10
采集批次: "2026年6月13日8点30分29秒"
采集批次ID: "20260613-083029-484"
去重键: "https://www.nvidia.com/en-gb/data-center/gb300-nvl72"
---

# NVIDIA GB300 NVL72
Built for the age of AI reasoning.

## Available Now

[ Read Press Release ](https://nvidianews.nvidia.com/news/nvidia-blackwell-ultra-ai-factory-platform-paves-way-for-age-of-ai-reasoning)

* Overview * Performance * Features * Specs * Resources * Get Started * Overview * Performance * Features * Specs * Resources * Get Started * Overview * Performance * Features * Specs * Resources * Get Started [ Read Press Release ](https://nvidianews.nvidia.com/news/nvidia-blackwell-ultra-ai-factory-platform-paves-way-for-age-of-ai-reasoning)

Overview

## Designed for AI Reasoning Performance

The NVIDIA GB300 NVL72 features a fully liquid-cooled, rack-scale architecture that integrates 72 [NVIDIA Blackwell Ultra GPUs](https://www.nvidia.com/en-gb/data-center/technologies/blackwell-architecture/) and 36 Arm®-based [NVIDIA Grace™ CPUs](https://www.nvidia.com/en-gb/data-center/grace-cpu/) into a single platform. The system delivers 1.5x more dense FP4 Tensor Core FLOPS and 2x higher attention performance compared to NVIDIA Blackwell GPUs. It’s purpose-built for test-time scaling inference and AI reasoning tasks. [AI factories](https://www.nvidia.com/en-gb/solutions/ai-factories/) accelerated by the GB300 NVL72—leveraging NVIDIA Quantum-X800 InfiniBand or Spectrum-X™ Ethernet, ConnectX-8 SuperNIC™s, and NVIDIA Mission Control management—deliver up to a 50x overall increase in AI factory output performance compared to [NVIDIA Hopper-based platforms](https://www.nvidia.com/en-gb/data-center/technologies/hopper-architecture/).

## Performance

## Scaling AI Factories to Unprecedented Levels

DeepSeek R1 ISL = 32K, OSL = 8K, GB300 NVL72 with FP4 Dynamo disaggregation. H100 with FP8 in-flight batching. Projected performance subject to change.

Experience next-level AI reasoning performance with the NVIDIA GB300 NVL72 platform. Compared to Hopper, the GB300 NVL72 delivers an impressive 10x boost in user responsiveness (TPS per user) and a 5x improvement in throughput (TPS per megawatt (MW)). Together, these advancements translate into a remarkable 50x leap in overall AI factory output.

Features

## Technological Breakthroughs

### AI Reasoning Inference

Test-time scaling and AI reasoning increase the compute necessary to achieve quality of service and maximum throughput. NVIDIA Blackwell Ultra’s [Tensor Cores](https://www.nvidia.com/en-gb/data-center/tensor-cores/) are supercharged with 2x the attention-layer acceleration and 1.5x more AI compute floating-point operations per second (FLOPS) compared to NVIDIA Blackwell GPUs.

### 288 GB of HBM3e

Larger memory capacity allows for larger batch sizing and maximum throughput performance. NVIDIA Blackwell Ultra GPU’s offer 1.5x larger HBM3e memory in combination with added AI compute, boosting AI reasoning throughput for the largest context lengths.

### NVIDIA Blackwell Architecture

The NVIDIA Blackwell architecture delivers groundbreaking advancements in accelerated computing, powering a new era of unparalleled performance, efficiency, and scale.

### NVIDIA ConnectX-8 SuperNIC

The [NVIDIA ConnectX-8 SuperNIC’s](https://resources.nvidia.com/en-us-accelerated-networking-resource-library/connectx-datasheet-c) input/output (IO) module hosts two ConnectX-8 devices, providing 800 gigabits per second (Gb/s) of network connectivity for each GPU in the NVIDIA GB300 NVL72. This delivers best-in-class remote direct-memory access (RDMA) capabilities with either [NVIDIA Quantum-X800 InfiniBand](https://www.nvidia.com/en-gb/networking/products/infiniband/quantum-x800/) or [Spectrum-X Ethernet](https://www.nvidia.com/en-gb/networking/spectrumx/) networking platforms, enabling peak AI workload efficiency.

### NVIDIA Grace CPU

The NVIDIA Grace CPU is a breakthrough processor designed for modern data center workloads. It provides outstanding performance and memory bandwidth with 2x the energy efficiency of today’s leading server processors.

### Fifth-Generation NVIDIA NVLink

Unlocking the full potential of accelerated computing requires seamless communication between every GPU. The fifth-generation of [NVIDIA NVLink™](https://www.nvidia.com/en-gb/data-center/nvlink/) is a scale–up interconnect that unleashes accelerated performance for AI reasoning models..

## NVIDIA Mission Control

NVIDIA Mission Control streamlines AI factory operations, from workloads to infrastructure, with world-class expertise delivered as software. It powers NVIDIA Grace Blackwell data centers, bringing instant agility for inference and training while providing full-stack intelligence for infrastructure resilience. Every enterprise can run AI with hyperscale efficiency, simplifying and accelerating AI experimentation.

[ Run Models, Automate the Essentials ](https://www.nvidia.com/en-gb/data-center/mission-control/)

## NVIDIA DGX Station

Develop and run cutting‑edge AI locally with NVIDIA DGX Station™, powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip. Accelerate large-scale training and inferencing workloads and easily scale them from deskside to data center.

[ Learn More About NVIDIA DGX Station ](https://www.nvidia.com/en-gb/products/workstations/dgx-station/)

Specifications

## NVIDIA GB300 NVL72¹

Configuration | 72 NVIDIA Blackwell Ultra GPUs, 36 NVIDIA Grace CPUs
---|---
NVLink Bandwidth | 130 TB/s
Fast Memory | 37 TB GPU Memory | Bandwidth | 20 TB | Up to 576 TB/s
CPU Memory | Bandwidth | 17 TB LPDDR5X | 14 TB/s
CPU Core Count | 2,592 Arm Neoverse V2 cores
FP4 Tensor Core | 1440 | 10802 PFLOPS
FP8/FP6 Tensor Core | 720 PFLOPS
INT8 Tensor Core | 24 POPS
FP16/BF16 Tensor Core | 360 PFLOPS
TF32 Tensor Core | 180 PFLOPS
FP32 | 6 PFLOPS
FP64 / FP64 Tensor Core | 100 TFLOPS

1\\. All Tensor Core specifications are with sparsity unless otherwise noted. 2\\. Without sparsity.

Resources

## Learn More About Grace Blackwell

1. Blogs
2. Sessions

[![NVIDIA Achieves Leading Agentic Coding Performance on First Agentic AI Benchmark](https://developer-blogs.nvidia.com/wp-content/uploads/2026/06/ai-agent-1-768x432.png)](https://developer.nvidia.com/blog/nvidia-achieves-leading-agentic-coding-performance-on-first-agentic-ai-benchmark/)
[ June 12, 2026 NVIDIA Achieves Leading Agentic Coding Performance on First Agentic AI Benchmark ](https://developer.nvidia.com/blog/nvidia-achieves-leading-agentic-coding-performance-on-first-agen…