By Nick Flaherty www.flaherty.co.uk
Server designs are changing, bringing more challenges to the embedded system designer.
Artesyn Embedded Technologies for example, has developed a new hyperscale media acceleration server sled based on the CG-OpenRack-19 architecture, which is inspired by rack-scale architectures.
The new CG19-GPU sled will be used by carriers to add NVIDIA Tesla P4 GPU processing to their network infrastructure systems. Optimised support for GPU and media processing workloads enables carriers to enhance their networks with improved video streaming and services that use video, such as digital advertising and augmented reality. Carriers also benefit from dramatically increased performance density and the ability to deploy new applications such as advanced analytics.
CG-OpenRack-19 is an open-source specification for scalable carrier-grade rack-level systems that integrate high-performance compute, storage and networking in a standard rack. It is designed to enable applications to be deployed quickly on off-the-shelf servers and storage systems.
CG-OpenRack-19 is an open-source specification for scalable carrier-grade rack-level systems that integrate high-performance compute, storage and networking in a standard rack. It is designed to enable applications to be deployed quickly on off-the-shelf servers and storage systems.
“Carrier networks need to rapidly deploy new applications on datacenter infrastructure, and open source hardware standards such as CG-OpenRack-19 are critical to this network transformation," said Linsey Miller, vice president of marketing for Artesyn Embedded Technologies. "We are bringing our deep knowledge of how to apply off-the-shelf technology along with our breadth of product and third-party ecosystem to solve today’s application deployment challenges, so that service providers can buy with confidence.”
“NVIDIA lets companies like Artesyn, and its telecom customers, deploy a GPU computing platform that powers AI applications capable of creating additional growth opportunities,” said Craig Weinstein, vice president of the America's Partner Organization at NVIDIA. “The Tesla P4 hyperscale GPU enables superior carrier services including, virtual reality, augmented reality and improved video streaming.”
Related rack and NVIDIA stories on the Embedded blog:
Microsoft Azure backs ARM chips for servers
King Cobra rugged server design provides a “total rack in 2U" ..
Arrow invests in Swindon for Artesyn configurable power supplies...
No comments:
Post a Comment