跳转到主要内容
Supermicro NGC-Ready Solutions

Designed for accelerated application development and deployment with scalable and agile infrastructure

Supermicro NGC-Ready Systems are validated for functionality and performance of AI software from NVIDIA NGC. These systems, together with NVIDIA NGC, enable customers to develop and deploy end-to-end AI solutions. The systems come pre-installed with operating system, container, CUDA environment necessary to run NVIDIA NGC software. Supermicro provides NVIDIA NGC software installation as an additional service. Enterprise-grade support for hardware and the operating system is available to help system administrators minimize system downtime, giving users the confidence to support AI workloads.

Powered by NVIDIA V100 and T4, the Supermicro NGC-Ready systems provide speedups for both training and inference. The NVIDIA NGC software is delivered as Docker containers, so they can run anywhere the Supermicro NGC-Ready Systems are deployed:

  • in data centers or cloud
  • in edge micro-datacenters
  • in distributed remote locations as environment-resilient and secured EGX servers

Backed by 26 years of experience delivering state-of-the art computing solutions, Supermicro systems are the most power-efficient, highest-performing, and highest-value server solutions on the market today. With support for ultra-fast networking and storage, as well as NVIDIA V100 and T4 GPUs, the Supermicro NGC-Ready systems are the most scalable and reliable servers to support AI. Customers can run their AI infrastructures with confidence.

NVIDIA NGC

NGC is NVIDIA’s hub for GPU-accelerated AI, ML, and HPC applications containers, SDKs and tools that provides faster access to performance-optimized software.

NGC container registry hosts containers for the top AI and data science software — all tuned, tested and optimized by NVIDIA — as well as fully tested containers for HPC applications and data analytics. These containers take advantage of NVIDIA GPUs on-premises and in the cloud. Each is fully optimized and works across a wide variety of Supermicro server solutions with NVIDIA GPUs.

NVIDIA NGC also hosts pre-trained models to help data scientists build high- accuracy models faster and offers industry-specific SDKs that simplify developing end-to-end AI solutions with training and inference.

NVIDIA is the leader in high performance GPU accelerators and software for AI software development and deployment. By bringing technologies from the latest AI research into NGC, customers can feel confident they have the best choices to develop their AI applications. With pre-trained AI models and transfer learning, customers can develop AI solutions much more rapidly than before by using NGC.

Solution Diagram

Supermicro NGC-Ready System Advantages

Supermicro NGC-Ready systems are validated for performance and functionality to run NGC containers. Combined with the NVIDIA NGC software, the high-end NGC-Ready systems can aggregate GPUs over fast network and storage to train big AI models with large data batches. NGC-Ready systems for Edge deploy AI applications for AI inferencing.

Supermicro NGC-Ready systems also offer the following capabilities and services:

  • Performance validated systems accelerate time to solution and allow IT managers to choose the right system for the appropriate AI workloads
  • Supermicro offers an optional service to enable Immediate AI software development when the systems are powered on. Operating System, CUDA, CUDA-X, NVIDIA drivers, and container infrastructure are preloaded
  • Access to the latest, cutting-edge deep learning AI models from NVIDIA NGC. Using pre-trained models and your own data, new AI models can be constructed efficiently
  • The NGC containers run anywhere, whether on the Supermicro systems in the data center, in edge micro datacenters, on edge servers, and in the cloud if cloud bursting is needed. Kubernetes automate deployment, scaling and management of containers in your production environment
  • Hardware, OS, and NVIDIA NGC support services are available
  • Service to setup rack, servers, storage, and network
  • Kubernetes consulting service for multi-server deployment

Supermicro offers NGC-Ready Systems as part of our AI/DL solution stack, including enterprise level support. Please speak to your Supermicro sales representative for more information.

Supermicro and NGC – Solution Architecture

Supermicro NGC Ready Systems support customers to run their entire AI infrastructure, from large data training to data center inference, to edge inference.

NGC Solution Architecture

Supermicro NGC-Ready Systems

Supermicro Validated NVIDIA GPU Cloud (NGC) Servers. Please check here for the latest published list.

SYS-4029GP-TVRTSYS-2029GP-TRSYS-2029GP-TRSYS-1019D-FHN13TPSYS-E403-9P-FN2TSYS-5019D-FN8TP
TrainingInferencing
DatacenterEdge
SYS-4029GP-TVRTSYS-2029GP-TRSYS-2029GP-TRSYS-1019D-FHN13TPSYS-E403-9P-FN2TSYS-5019D-FN8TP
CPU2 * Xeon Scalable2 * Xeon Scalable2 * Xeon Scalable1 * Xeon D1 * Xeon D1 * Xeon D
Accelerator8 * V100 NVLink2 * V1004 * T42 * T42 * T41 * T4
Memory6TB4TB4TB512GB512GB512GB
Storage16 * SATA3; 8 * NVMe8 * SATA3; 2 * NVMe8 * SATA3; 2 * NVMe4 * SATA34 * SATA34 * SATA3
Network100GbE, GPU Direct RDMA100GbE, GPU Direct RDMA100GbE, GPU Direct RDMA10GbE, 10GSFP+10GbE, 10GSFP+10GBase-T, 10GSFP+
Form Factor4U Rackmount2U Rackmount2U RackmountShort DepthBox PC, Short DepthWall mount