Advanced Computing in the Age of AI | Sunday, April 28, 2024

Inspur Announces Its New G7 Server Platform Supports Latest 4th Gen Intel Xeon Scalable Processors 

SAN JOSE, Calif., Jan. 10, 2023 -- Inspur Information, a leading IT infrastructure solutions provider, announced that its G7 server platform fully supports 4th Gen Intel Xeon Scalable processors. The 16 different types of servers making up the brand-new lineup are industry-leading in terms of performance, openness, intelligent management, and sustainability. Compared with the previous generation of Intel-based products, performance is improved by up to 61% and the computing performance per unit of power consumption is improved by up to 30%. The server platform is designed to be deployed in general-purpose computing, critical computing, AI, and other application scenarios.

Inspur Information’s brand-new G7 platform was designed around the ethos of green technology, open-source solutions, precise engineering, and intelligent management. It is a leading example in the industry in terms of system design, energy savings, and operation & maintenance management. G7 servers support diversified computing, with the most comprehensive product lineup in the industry. With its focus in green energy, it supports cold plate and immersion cooling schemes, and has unique cooling designs such as T-shaped radiator and advance heat detection with intelligent regulation, all working together to reduce energy consumption by up to 30%. Cloud operation and maintenance is also supported for intelligent fault diagnosis with an accuracy rate up to 95%.

Green Technology for Better TCO

The G7 server platform fully supports cold-plate and full-immersion liquid cooling, and can be delivered as a rack-level liquid cooling solution. This provides a significant reduction in PUE without sacrificing performance, maintainability, sustainability, or energy-savings, resulting in a substantial reduction in TCO.

When equipped for air-cooled heat dissipation, the G7 server platform utilizes unique design elements including component-level intelligent control and precision temperature detection, a T-shaped radiator, and an optimized air outlet channel with temperature monitoring. These features work in tandem to reduce power consumption and improve heat dissipation efficiency. Fans are capable of providing the necessary volume of air according to the workload. An air-cooled G7 server can support two 350W CPUs in a 1U space, achieving 30%+ cost reduction in terms of energy savings compared to the previous generation.

Intelligent control technology can automatically monitor resource utilization, dynamically adjust the operating status of each component of the server, supply power according to workload needs, and achieve data center TCO optimization.

In terms of environmental protection, G7 servers are green and sustainable due to conscientious choices in both internal materials and packaging. Its internal electronic materials meet ROHS standards (no lead, cadmium and other 4 types of harmful metals). The packaging materials use 100% recyclable materials, reducing the use of plastic materials by 30%. If this same level of green initiative was applied to all servers around the globe, more than 94,000 tons of plastic would be saved.

Open Computing for Lower-Cost Deployment and Innovation Acceleration

The G7 server platform has upgraded its level of integration with open-source software and open architecture for improved development and deployment capabilities.

The G7 server platform integrates open software protocols, such as OpenBMC, Redfish, and CXL. Redfish is compatible with a wide range of communication protocols, and defines the URLs of different interfaces by covering different function modules, enabling server management via modularization. OpenBMC offers system management capabilities across heterogeneous systems, and offers in-depth and flexible customization of services.

Adopting the DC-SCM open standardized management module boosts server security management capabilities, decoupling BMC and ROT development cycles from CPU iterations, which accelerates updates of the security management chip and CPU, allowing better cross-platform compatibility.

Precision Engineering to Ensure Business Continuity

Data center server security is fundamental to routine business operations. Therefore, the G7 server platform adopts several mechanisms to ensure the reliability and stability of business operations with precision engineering across multiple levels, such as hardware, firmware, and system. For example, at the firmware level, the BIOS/BMC redundancy design enables secure and smooth firmware batch upgrades in data centers to ensure service continuity.

Today, there are increasing needs for intelligent network interface cards (NIC) being installed in servers, and the power consumption of intelligent NICs is increasing. To ensure a secure power supply after an intelligent NIC is upgraded, Inspur Information has redefined power supply standards to enable on-demand power output on the mainboard and enable both the intelligent NIC and the processor to start at the same time, which fundamentally resolves the power supply problem for intelligent NIC in bare-metal scenarios.

Intelligent Management to Provide Ease of Use

The G7 server platform supports cloud-based operation & maintenance, and can utilize online intelligent diagnosis of faults with an accuracy rate of up to 95%. Through the intelligent power consumption management function, the power consumption of stand-alone machines and data centers can be dynamically managed, which can save energy consumption by approximately 15%.

In terms of hardware operation and maintenance, the server uses different colors to mark hot-swappable components and has dynamic workload status indicator lights and front-mounted type-C management ports that can be connected to mobile phones to improve operation and maintenance efficiency.

The whole product line supports tool-free operation and maintenance. The front-mounted I/O outlet design is added to improve the space utilization rate of the equipment room. The temperature at the front of the cabinet is kept at approximately 25℃, extending the life of heat-sensitive components such as smart network cards by more than threefold.

G7 server family equipped with the latest 4th Gen Intel Xeon Scalable processors:

  • TS860G7 is a high-end server optimized for critical applications that is based on 4th Gen Intel Xeon Scalable processor. It can support a maximum of eight sockets in a 6U space. The system integrates 128 DIMM slots of DDR5 memory to achieve the highest computing density in the industry. It is oriented to large-scale core databases, SAP HANA memory computing, CSP cloud applications, and other scenarios.
  • NF8260G7 is a high-density server for cloud-optimized scenarios like CSP cloud applications. The server’s CPUs contain up to 60 cores. It supports various storage configurations, up to a 100 Gbps network, and a 4x 2700W redundant power supply.
  • NF5180G7 is a general-purpose density-optimized rack server, designed to maximize performance, density, and scalability in 1U space. It has three cooling solutions: air cooling, cold-plate liquid cooling, and immersion liquid cooling, making it suitable for high-performance computing, virtualization, and other computing-intensive application scenarios.
  • NF5280G7 is full-fledged rack server offering strong computing performance in a 2U space. It has comprehensive ecosystem compatibility, and high configuration flexibility to meet the application scenario requirements of various industries.
  • i24G7 is a high-density computing multi-node server. It can provide four dual-socket server nodes in 2U space. Each node supports up to two sockets of 4th Gen Intel Xeon Scalable processors, and 16 DIMM slots of DDR5 memory. It is designed for high-performance, virtualization, and other computing-intensive application scenarios.
  • NF5266G7 is a high-density storage server that adopts an innovative three-tier storage architecture in a 2U space to achieve the ultimate balance of computing and storage I/O. Its computing performance is up to 60% higher than the previous generation. It is suitable for scenarios such as big data, CDN, hyper-convergence, and distributed storage.
  • ORS3000S is a rack-scale system suitable for centralized power supplies and general-purpose rack servers. It optimizes the power consumption of nodes and refrigeration equipment, supports liquid cooling, and adopts an optimized and flexible network modular design to adapt to various Internet and enterprise applications.
  • NF5468G7 is a 4U versatile AI computing platform, powered by 2x 4th Gen Intel Xeon Scalable processors with up to 112 cores and a 700W TDP. It can be configured to handle a multitude of specialized AI tasks including deep learning, metaverse generation, AIGC, and AI for Science by using an array advanced and specialized accelerators.
  • NF5688G7 is a 6U hyperscale AI training platform, featuring industry-leading performance, ultimate I/O expansion and ultra-high energy efficiency. Equipped with 8x 700W GPUs, it is built to handle the most demanding AI computing tasks like trillion-parameter AI model training, massive recommendation systems, and metaverse workloads.

To learn more about the G7 server platform family equipped with the latest 4th Gen Intel Xeon Scalable Processors, click here.

About Inspur Information

Inspur Information is a leading provider of data center infrastructure, cloud computing, and AI solutions. It is the world’s 2nd largest server manufacturer. Through engineering and innovation, Inspur Information delivers cutting-edge computing hardware design and extensive product offerings to address important technology sectors such as open computing, cloud data center, AI, and deep learning. Performance-optimized and purpose-built, our world-class solutions empower customers to tackle specific workloads and real-world challenges.


Source: Inspur Information

EnterpriseAI