Home > Events > GTC2020

GTC 2020 Inspur Technical Sessions

Please click here for more GTC 2020 sessions.

GTC Digital is the great training, research, insights, and direct access to the brilliant minds of NVIDIA’s GPU Technology Conference, now online.
Join Inspur and supercharge your data center with Inspur full-stack AI solutions.

Meet Us Here


Technical speaking sessions to introduce Inspur's Full-Stack AI Capabilities and Inspur's GPU Server Design.


Inspur offers a comprehensive full-stack of AI capabilities that spans the full range of application, framework, management and hardware platforms

Sessions Overview

Alan Chang
Deputy GM & Sr. Director of Server Product Line
Extreme Design of AI Server:

In the era of intelligent computing, Inspur is one of the world’s leading hardware innovators and solutions providers for AI and deep learning applications. Along with the rapid development of the AI innovation, the three main driving forces, Algorithm, Big Data and Computing have much more strict requirement for meeting the customers’ scenarios demand. Facing with the fast-growing and increasingly complex deep neural networks, how can we overcome the challenges of the AI Infrastructure R&D and deployment? Inspur Extreme Design of the AI Server can solve your worries.

Vangel Bojaxhi
Global AI&HPC Director
Inspur Full-Stack AI Solutions:

In the era of intelligent computing, Inspur is one of the world’s leading hardware innovators and solutions providers for AI and deep learning applications. Along with the rapid increasing trend of global AI market, we can foresee that in the next 2 to 5 years, a large number of AI technologies will move from the innovation phase to the production phase. Facing with the great chances of AI industry, how to provide customers with optimized computing infrastructure as well as algorithm and resources solution for application deployment? Inspur is developing newer, smarter end-to-end solutions that help businesses leverage and tackle intelligent technologies in the data center.


Leading AI Solution


Powerful Scale-up AI Super Server

16*NVSwitch-Enabled V100 GPU, Unprecedented 2 PetaFLOPS in One Server

  • Extreme Performance 16 NVSwitch-enabled V100 GPUs to offer unbeatable 2 petaFLOPS AI computing performance, and tackle the toughest AI&HPC challenges
  • Extreme Connectivity Based on HGX-2 platform, it adopts NVSwitch interconnect fabric, enabling 48-channel, 2.4TB/s full-chip cluster high-speed interconnects
  • Extreme Throughout 512GB HBM2 globally shared ultra-fast graphics cache, and provides nearly linear AI computing performance scaling
  • Flexible GPU topology Free to be configured with different topologies for various applications


Industry 1st 4U 8GPU NVSwitch-Empowered AI Server

Unmatchable Performance and Flexibility for Larger Scale Model Training

  • 8 NVSwitch-empowered V100 GPUs in 4U, deliver 1 petaFLOPS of superb performance
  • Enable faster AI training with lower cost compared to more servers with fewer GPUs each
  • 6KW in 4U design, make it easy to deploy in power-constrained racks
  • Designed for a wide range of deep learning and HPC applications


Extreme High Density AI Server

8*NVLink-Enabled V100 GPU in 2U Compact Form Factor

  • 8 NVLink-empowered V100 GPUs in a compact 2U design, deliver uncompromising performance in maximum density
  • Options of air cooling or air-liquid hybrid cooling, make it easy to deploy in Green Datacenters with lower PUE
  • Power a variety of AI&HPC applications with flexible configurations


Popular Inference Server with Incredible Flexibility

Popular Inference Server with 16*T4 GPU in 4U

  • Up to 16 T4 GPUs in one server, turbocharge AI Inference to gain real-time insight
  • Up to 8 NVLink-empowered V100 GPUs, supercharge large-scale training of AI models
  • Provide perfect combination of performance and internal storage with 384 TB
  • One click to change GPU topology with BMC, drive a broad range of applications from AI cloud, FinTech, telecom to healthcare


Flexible Configurations for Various Application Scenarios

Application scenario: cloud computing, big data, and deep learning applications of hyper-scale data centers.

Product Highlights:

  • Extreme Performance: Strong compute power, full flash configuration and zero storage bottleneck.
  • Flexible Configuration:Full modular design and over 30 application-based configuration options. On-demand combination of storage, IO and heterogenous modules.
  • Smart and Open Management: Optimized for large-size data center (24/7 operation monitoring, intelligent fault isolation and troubleshooting and open management protocols).

AI Station

A complete AI deep learning cluster management software enabling simple and flexible deep learning, and makes it easy to train and combine popular model types across multiple GPUs and servers.

Auto ML

The world’s first highly parallel extension to provide on-premise and cloud deployment, and it enables non-professionals to build network models with minimal operations