Inspur Information MetaEngine announces support for next-generation NVIDIA L40 GPUs and the latest Omniverse software stack

SAN JOSE, California, September 21, 2022 – Inspur Information announced that its NVIDIA OVX solution MetaEngine now supports NVIDIA L40 GPUs and the latest Omniverse Enterprise software stack to help users create digital assets more efficiently. Inspur Information also announced its virtual human and digital twin solutions that will enable a wide range of Metaverse applications by simplifying and speeding up the creation process.

MetaEngine is the underlying computational foundation of Inspur Information for the Metaverse. It provides the foundational technologies and tools needed to build and run the Metaverse, and supports a variety of workloads including AI, rendering, simulation, and modeling. This power and versatility allows MetaEngine to meet the diverse computational needs of the four main pillars of Metaverse creation: collaborative creation, real-time rendering, high-precision simulation, and intelligent interaction. Thanks to its fast, non-blocking network channel, it can also be expanded into large computing clusters.

A single MetaEngine server supports 8x NVIDIA L40 GPUs and full-link PCIe Gen5 for a 2x performance boost. It also supports the NVIDIA ConnectX–7 network adapter, doubling the bandwidth to and from the server. Powered by NVIDIA Omniverse Enterprise, it can perfectly combine the graphics and AI computation of high-performance GPUs with high-speed memory access and low-latency networking. A single MetaEngine server now supports 4000 AIGC digital scenes per second, doubling the previous version. It improves the real-time rendering speed of digital twin scenes by 50% and improves the performance of virtual human language training by 1.6 times. 32 MetaEngine servers can be combined into a scalable cluster unit with multiple scalable units, which can then be scaled out to form an even larger cluster for exceptional computing power and ultra-high network bandwidth.

MetaEngine supports NVIDIA Omniverse Enterprise for more efficient creation of virtual humans and digital twins. When Inspur’s Yuan 1.0 runs on MetaEngine’s NVIDIA OVX solution, users can create virtual humans that can see, hear, understand, and communicate with humans. The extensive physical simulation library for digital twins allows the 3D world to better mimic the laws of physics and supports the connection of more than 20 types of 3D tools to create 3D assets together. In the future, neural networks can be used to speed up the automatic reconstruction of 2D images into 3D models.

MetaEngine integrates rich ecosystems. The integrated, enterprise-level management and orchestration software provides users with an extensive suite of modeling, rendering, and simulation tools. Inspur’s AIStation resource platform, algorithm platform and application optimization service enable a large ecosystem. It can support a variety of metaverse applications for different scenarios, accelerate the construction of digital twin worlds through collaborative optimization, and create an efficient collaborative metaverse development experience for users.

The MetaEngine-based Virtual Human and Digital Twin solutions were also introduced. The Virtual Human solution integrates the complete technology and tools required for the construction and operation of virtual humans, providing powerful computational, software and AI support for the creation of large-scale, highly complex and highly realistic virtual human scenes and the Realtime Rendering Simulation . The digital twin solution provides a highly realistic simulation environment that integrates AI-driven, high-precision simulation algorithms and creates an application closed-loop for the virtual-real fusion of digital twins. The Virtual Human and Digital Twin solutions fully integrate the capabilities of NVIDIA Omniverse, providing a platform for hardware and software co-optimization that enables users to produce virtual human animations, true multi-person, multi-tool collaborative models to accelerate physical simulation and other metaverse technologies, greatly simplifying the construction of metaverse applications.

As a leading manufacturer of AI computing, Inspur Information will continue to drive the upgrade and iteration of MetaEngine, work with partners to create an integrated Metaverse solution, and develop scenario-specific applications for industries such as architecture, engineering, construction, operations, game development , manufacturing, media and entertainment.

About Inspur information

Inspur Information is a leading provider of data center infrastructure, cloud computing and AI solutions. It is the second largest server manufacturer in the world. Through engineering and innovation, Inspur Information delivers cutting-edge computing hardware design and extensive product offerings in key technology areas such as open computing, cloud data centers, AI and deep learning. Performance-optimized and purpose-built, our best-in-class solutions enable customers to solve specific workloads and real-world challenges. To learn more, visit https://www.inspursystems.com.


Source: Inspur

Laisser un commentaire

Votre adresse e-mail ne sera pas publiée.