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

Accelerate the largest AI, HPC, cloud, and hyperscale workloads.

AI models are exploding in complexity and size as they enhance deep recommender systems containing tens of terabytes of data, improve conversational AI with hundreds of billions of parameters, and enable scientific discoveries. Scaling these massive models requires new architectures with fast access to a large pool of memory and a tight coupling of the CPU and GPU. The NVIDIA Grace™ CPU delivers high performance, power efficiency, and high-bandwidth connectivity that can be used in diverse configurations for different data center needs. 


Turbocharging Recommender System Accuracy with the NVIDIA Grace Hopper Superchip.

The Developer Conference for the Era of AI and the Metaverse

Conference & Training September 19 - 22 | Keynote September 20

Join us this September for a GTC that will inspire your next big idea. This is a don’t-miss opportunity to learn from experts and leaders in their fields on how AI is transforming industries and profoundly impacting the world. It all happens online September 19–22.

  • Inside Grace

    • Jonathon Evans, NVIDIA

    This talk will give an overview of the NVIDIA Grace™ server system on chip (SOC).

  • Pushing the Frontier of AI Performance and Efficiency with Modern Infrastructure

    • Bhumik Patel, Arm
    • David Lecomber, Arm

    As the adoption of Arm® Neoverse-based computing solutions grows across cloud, 5G, HPC, and edge markets, so does the availability of software enabled and optimized for the Arm ecosystem. We'll review the current state of Arm libraries and developer tools available for optimizing HPC, AI, and other infrastructure codes to run on Arm.

  • A Deep Dive into the Latest HPC Software

    • Timothy Costa, NVIDIA

    Take a deep dive into the latest developments in NVIDIA software for HPC applications, including a comprehensive look at what’s new in programming models, compilers, libraries, and tools. We'll cover topics of interest to HPC developers, targeting traditional HPC modeling and simulation, HPC+AI, scientific visualization, and quantum computing.

Speakers at GTC22

Take a look at the Grace lineup.

NVIDIA Grace Hopper Superchip

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

NVIDIA Grace Hopper Superchip
NVIDIA Grace CPU Superchip

NVIDIA Grace CPU Superchip

The NVIDIA Grace CPU Superchip uses the NVLink-C2C technology to deliver 144 Arm® Neoverse V2 cores and 1 terabyte per second (TB/s) of memory bandwidth.

Explore Grace reference designs for modern data center workloads.

System designs for digital twins, AI, and high-performance computing.

OVX-Digital Twins & Omniverse


for digital twins and NVIDIA Omniverse™.

NVIDIA Grace CPU Superchip
NVIDIA BlueField®-3



for HPC.

NVIDIA Grace CPU Superchip
NVIDIA BlueField-3
OEM-defined input/output (IO)

HGX-AI Training, Inference & HPC


for AI training, inference, and HPC.

NVIDIA Grace Hopper Superchip
NVIDIA BlueField-3
OEM-defined IO / fourth-generation NVLink

Designed to Solve Complex Problems

The NVIDIA Grace CPU leverages the flexibility of the Arm® architecture to create a CPU and server architecture designed from the ground up for accelerated computing. This innovative design will deliver up to 30X higher aggregate bandwidth compared to today's fastest servers and up to 10X higher performance for applications running terabytes of data. NVIDIA Grace is designed to enable scientists and researchers to train the world’s largest models to solve the most complex problems.

Learn more about the latest technical innovations.

Accelerate CPU-to-GPU connections with NVLink-C2C.

Solving the largest AI and HPC problems requires high-capacity and high-bandwidth memory (HBM). The fourth-generation NVIDIA NVLink-C2C delivers 900 gigabytes per second (GB/s) of bidirectional bandwidth between the NVIDIA Grace CPU and NVIDIA GPUs. The connection provides a unified, cache-coherent memory address space that combines system and HBM GPU memory for simplified programmability. This coherent, high-bandwidth connection between CPU and GPUs is key to accelerating tomorrow’s most complex problems.

Harness high-bandwidth memory with LPDDR5X.

NVIDIA Grace is the first server CPU to harness LPDDR5X memory with server-class reliability through mechanisms like error-correcting code (ECC) to meet the demands of the data center, while delivering 2X the memory bandwidth and up to 10X better energy efficiency compared to today’s server memory. The LPDDR5X solution coupled with NVIDIA Grace’s large high-performance last-level cache delivers the bandwidth necessary for large models while reducing system power to maximize performance for next-generation workloads. 

Boost performance and efficiency with Arm Neoverse V2 cores.

As the parallel compute capabilities of GPUs continue to advance, workloads can still be gated by serial tasks run on the CPU. A fast and efficient CPU is a critical component of system design to enable maximum workload acceleration. The NVIDIA Grace CPU integrates Arm Neoverse V2 cores to deliver high performance in a power-efficient design, making it easier for scientists and researchers to do their life’s work.

Meet our partners.

Atos logo
DELL logo
Fii logo
Gigabyte logo
H3c logo
Hewlett Packard Enterprise Logo
Inspur logo
Lenovo logo
Nettrix logo

Turbocharging Recommender System Accuracy with the NVIDIA Grace Hopper Superchip.