Introducing Grace

5 min read Original article ↗

Cloud and Data Center Solutions

NVIDIA Grace

Purpose-built to solve the world’s largest computing problems.

Introduction

The Engine of the Next-Generation Data Center

As models explode in complexity, accelerated computing, and energy efficiency are becoming critical to meet the demands of AI. The NVIDIA Grace™ CPU is a groundbreaking Arm® CPU with uncompromising performance and efficiency. It can be tightly coupled with a GPU to supercharge accelerated computing or deployed as a powerful, efficient standalone CPU. The NVIDIA Grace CPU is the foundation of next-generation data centers and can be used in diverse configurations for different data center needs.

NVIDIA Vera

The NVIDIA Vera CPU is purpose-built for the next generation of data centers—engineered to pair seamlessly with NVIDIA GPUs for Gigascale AI or operate independently to handle the analytics, orchestration, and storage workloads that generate intelligence and keep data moving. 

Benefits

Technological Breakthroughs

Accelerate CPU-to-GPU Connections With NVLink-C2C

NVIDIA NVLink™ Chip to Chip (C2C) offers 900 gigabytes per second (GB/s) of bidirectional bandwidth, 7x faster than PCIe Gen 5, enabling focus on development instead of memory management.

Tap Into Energy-Efficient CPU Memory With LPDDR5X

NVIDIA Grace uses LPDDR5X memory with error-correction code (ECC) for server-class reliability, while offering 5x better energy efficiency—ideal for cloud, enterprise, and high-performance computing (HPC) workloads.

Boost Performance and Efficiency With Arm Cores and NVIDIA SCF

NVIDIA Grace uses high-performance Arm Neoverse™ V2 cores coupled with NVIDIA Scalable Coherency Fabric (SCF) to maximize performance and energy efficiency for data center workloads.

Diverse Configurations for Accelerated and CPU Workloads

The NVIDIA Grace portfolio includes tightly coupled CPU and GPU systems, such as Grace Blackwell and Grace Hopper™ for accelerated computing, as well as CPU-only solutions with the Grace CPU.

NVIDIA Vera CPU

The NVIDIA Vera CPU revolutionizes the modern data center with an advanced architecture that features 88 NVIDIA-designed, high-performance, custom Arm® CPU cores and 1.2 terabytes per second of memory bandwidth.

Products

Explore the Grace Lineup

NVIDIA Grace CPU

Breakthrough CPU performance and efficiency for the modern data center.

NVIDIA Grace CPU Superchip

The Grace CPU Superchip is composed of two Grace CPUs connected coherently over NVIDIA NVLink-C2C at 900 GB/s. It packs 144 Arm Neoverse V2 cores into a single module, with server-class LPDDR5X memory that delivers up to 1 terabyte per second (TB/s) of memory bandwidth.

NVIDIA Grace CPU C1

The NVIDIA Grace CPU C1 is a single-socket, high-performance server platform that doesn't compromise performance or bandwidth. It’s optimized for scalable and edge platforms, including hyperscale cloud, content delivery network (CDN), storage, telco, and other high-performance edge platforms.

NVIDIA Grace Blackwell

Powering the new era of computing with Grace CPU and NVIDIA Blackwell GPU.

NVIDIA GB200 NVL72

The NVIDIA GB200 NVL72 connects 36 NVIDIA Grace CPUs and 72 NVIDIA Blackwell GPUs in a rack-scale design, supercharging generative AI, data processing, and high-performance computing. The NVIDIA GB200 NVL72 acts as a single, massive GPU and delivers 30x faster real-time inference for trillion-parameter large language models (LLMs).

NVIDIA GB300 NVL72

NVIDIA GB300 NVL72 features 35x more AI inference performance than Hopper platforms, 40TB of fast memory, and networking platform integration with NVIDIA ConnectX®-8 SuperNICs using Quantum-X800 InfiniBand or Spectrum™-X Ethernet. NVIDIA Blackwell Ultra delivers breakthrough performance on the most complex workloads, from agentic systems and reasoning to 30x faster real-time video generation.

NVIDIA GB200 NVL4

NVIDIA GB200 NVL4 unlocks the future of converged HPC and AI, delivering revolutionary performance through a bridge connecting four NVIDIA NVLink Blackwell GPUs unified with two Grace CPUs over NVLink-C2C interconnect. Compatible with liquid-cooled, NVIDIA MGX™ modular servers, it provides up to 2X performance for scientific computing, AI for science training, and inference applications over the prior generation.

NVIDIA Grace Hopper

Connecting the NVIDIA Grace CPU and Hopper for large-scale AI and HPC applications.

NVIDIA GH200 Grace Hopper Superchip

The NVIDIA Grace Hopper Superchip combines the Grace and Hopper architectures using NVLink-C2C to deliver a CPU+GPU coherent memory model for accelerated AI and HPC applications.

NVIDIA GH200 NVL2

The NVIDIA GH200 NVL2 fully connects two GH200 Superchips with NVLink, delivering up to 288 GB of high-bandwidth memory, 10 TB/s of memory bandwidth, and 1.2 TB of fast memory.

Introducing NVIDIA DGX Spark

DGX Spark brings the power of NVIDIA Grace Blackwell™ to developer desktops. The GB10 Superchip, combined with 128 GB of unified system memory, lets AI researchers, data scientists, and students work with AI models locally with up to 200 billion parameters.

Resources

Discover More About NVIDIA Grace

AI Factory for the New Industrial Revolution

Discover how the AI factory generates tokens to help build a future filled with endless possibilities—accelerated by human ingenuity and NVIDIA.

NVIDIA Grace CPU Superchip

The breakthrough CPU for the modern data center that delivers twice the performance at the same power as leading traditional CPUs.

GTC24 Keynote With NVIDIA CEO Jensen Huang

Watch NVIDIA CEO Jensen Huang’s GTC keynote to catch all the announcements on AI advances that are shaping our future.

Explore More Resources

Stay Up to Date on NVIDIA News

Sign up for the latest news, updates, and more from NVIDIA.