Inspur Blog

2021 Year in Review and Looking Ahead at 2022

Two years into the global pandemic, the world is finding its bearings and settling into a new normal. Businesses resumed, figuring their way around obstacles to ensure continued production and delivery of goods and services. Similarly, while intermittently beset by ongoing challenges like logistics and supply chain disruptions, Inspur Information presses on, continuing to develop new and innovative solutions in AI, cloud computing, and open computing to support various industries.

2021 Achievements and Milestones

In the last year, Inspur’s growth and productivity garnered several major recognitions. For the 10th consecutive quarter, Inspur ranked 1st in the public cloud infrastructure global market according to Synergy. The Forrester Wave report recently identified Inspur as a strong performer in AI Infrastructure, and according to Gartner’s latest data, Inspur Information is now the 2nd largest provider in the global server market.


Launch of the M6 Server Series based on next-generation Intel Xeon Scalable

Compute performance has been a driving force of the digital economy with the rise of online education, remote work, and intelligent manufacturing. Inspur introduced the new M6 series of 16 new servers, based on Intel’s 3rd generation Xeon Scalable processors to deliver industry-leading deployment density, performance, hardware decoupling, and reliability that improves performance by 46% compared to the previous generation.

Launch of 3 new AI servers supporting NVIDIA A30 and A10 GPUs at GTC Spring

The latest NVIDIA A30 and A10 GPUs provide customers the ability to meet the demands of today’s demanding AI and graphics applications. Inspur AI servers are among the first in the industry to fully support NVIDIA A30 and A10 GPUs, offering diverse product options to meet the needs of different sectors and scenarios.

AI upgrade of Northwestern Medicine’s data system

Using Inspur’s A100 training platform NF5488M-D, Northwestern Medicine implemented a high-performance data pipeline to enable deep learning experiments without the need to duplicate legacy health system data, which improved efficiency and dramatically reduced costs. Read the use case.


ASC20-21 Student Supercomputer Challenge

Inspur sponsored the annual ASC20-21 Student Supercomputer Challenge, the world’s largest Supercomputer Hackathon, to train and develop supercomputing talent in college-age youths. The competition attracted over 300 university teams from around the world with 28 teams advancing to the finals.


Accelerating DNA lab’s archaeological genome analysis

Inspur partnered with a laboratory studying ancient DNA to develop a gene sequencing acceleration program which sped up the identification and analysis of DNA samples by 39 times. Inspur’s solution also enabled AI technology platforms that parse small fragments of severely deteriorated DNA samples to improve accuracy. This partnership helps the DNA lab take a big step forward in unraveling the mysteries of ancient human civilizations. More details in this blog post.


Launch of liquid-cooled AI server with NVIDIA A100 GPUs at ISC High Performance Digital 2021

High-performance computing and AI applications are calling for higher energy-saving standards. The newly updated NF5488A5 reaches a GPU-to-GPU communication bandwidth of 600GB/s and effectively reduces the Power Usage Effectiveness (PUE) to 1.1, while maintaining high performance capabilities: it ranked 1st in 11 out of 16 tasks in the 2021 MLPerf Inference v1.0 closed division data center benchmark.


Aiding Asian elephant conservation with Inspur AI

In July a herd of Asian elephants strayed from their habitat and entered human settlements. Inspur quickly set up an emergency work group and conducted tracking of the elephants’ movements using AI powered systems to safeguard the safety of both local residents and elephants until the elephants returned to their natural habitat. Read more about Inspur’s efforts on Asian elephant protection here.

Empowering ElectrifAi’s livestock welfare solution

ElectrifAi, a provider of business-ready machine learning models, offers an animal welfare AI product AnimalAI that monitors animal wellness and harm reduction in the livestock industry. Using Inspur’s NF5280M5 AI server, AnimalAi is able to process data from cameras and run machine learning locally and efficiently with a continuous and reliable connection. Read the use case here.

Co-publishing the Open Computing White Paper with Omdia

Inspur and Omdia, a world-renowned market research organization specializing in technology, released the Open Compute White Paper to facilitate collaborative innovation in the open computing industry. Open computing is a brand-new model for industrial collaboration where participants can share products, specifications and IPs of IT infrastructure to accelerate the adoption of innovative data center technologies.

Hosting the 10th anniversary of OCP

Once a niche division in the hardware industry, open computing is now considered one of the most important drivers of innovation for future data centers. As a longstanding member of the OCP community, Inspur hosted OCP Tech Day to mark the 10th anniversary of the Open Compute Project (OCP) Foundation, gathering technology experts from 23 leading companies to share the initiative’s achievements over the past 10 years and discuss sustainability goals for the next decade. Read the blog for more details.


Launch of the Yuan 1.0 Language Model

Inspur Information’s AI Research division announced Yuan 1.0, one of the world’s largest language models with 245 billion parameters and 5 TB of datasets. Its parallel GPU computing and performance in zero-shot and few-shot learning gives it the ability to generate language content that is difficult to distinguish from human-generated content. Read more details in this press release.

Top results in MLPerf Inference V1.1

In the very competitive closed division of MLPerf Inference v1.1, Inspur came in 1st in 15 out of 30 tasks, scoring top results with its AI servers NF5488A5, NF5688M6, and NE5260M5.


Joint development of open storage solution with Samsung at OCP Summit

In collaboration to meet the hyperscale market’s need for high-density, high-performance storage, Inspur and Samsung launched the Poseidon V2 E3.x reference system, an open all-flash storage solution with flash memory capacity of up to 256TB. It can accommodate two rows of flash dies to support this larger capacity, while supporting hot plug functionality, lower power consumption, and heat dissipation requirements. Read more in the press release.

Joint development of ARM-based open server with Ampere at OCP Summit

Also announced at OCP Summit is an ARM-based OCP platform developed in conjunction with Ampere. The industry’s first aarch64-based open platform features two Ampere Altra Max ARM CPUs on Inspur’s 2U server design to provide users with flexible, scalable, and secure cloud native servers.

Announcement of full support for NVDIA’s inference platform at GTC Fall

Inspur announced that its AI and Edge inference servers will support NVIDIA A100 and A30 GPUs, along with the newly announced A2 Tensor Core GPU throughout its entire inference server portfolio. Inspur has launched a comprehensive inference product line built for applications from data centers to edge computing, providing high performance for users across various application scenarios.


Top scoring AI Performance in MLPerf Training v1.1

MLPerf is an AI performance benchmark that has become an essential reference for customers looking for AI solutions. In the latest MLPerf Training v1.1, 14 organizations participated in 180 closed division and 6 open division tasks. Inspur AI servers NF5488A5 and NF5688M6 topped 7 out of 8 single-node closed division tasks, rounding out a total of 44 MLPerf titles won by Inspur in 2021.

Supporting Foster Children to Pursue STEM

As a way to connect with and give back to the community, Inspur Information partnered with charity organization Together We Rise to assemble and donate gifts of STEM activity boxes for foster children in the area. For a year that was difficult for young learners, Inspur saw the importance of engaged, hands-on educational experiences, and access to STEM concepts. Read more at the blog here.

Onward to 2022

Currently in the works is the Meta Engine Lab, a world class research and testing lab to bring about the most cutting-edge innovations, highest quality servers and most optimized solutions for the next generation of computing. Looking ahead at the rest of the year and beyond, Inspur Information will continue to develop new technologies, break more records, and deliver better products to serve our customers, so they may tackle new challenges and achieve business success in the new digital economy.

Leave a Reply

Your email address will not be published. Required fields are marked *