Artificial Intelligence 6 min read

Alibaba Cloud Scientist Jiang Xiaowei Inducted into HPCA Hall of Fame for AI Server and CPU Innovations

Alibaba Cloud scientist Jiang Xiaowei, a pioneer in CPU architecture, chip design, networking, and virtualization, was recently inducted into the HPCA Hall of Fame, marking the first cloud‑computing scholar honored, and highlighting his contributions to the EFLOPS AI server cluster that dramatically speeds up large‑scale model training.

Alibaba Cloud Infrastructure
Alibaba Cloud Infrastructure
Alibaba Cloud Infrastructure
Alibaba Cloud Scientist Jiang Xiaowei Inducted into HPCA Hall of Fame for AI Server and CPU Innovations

Recently, the IEEE-sponsored International Symposium on High Performance Computer Architecture (HPCA) announced its latest Hall of Fame inductees, with Alibaba Cloud Infrastructure Server R&D team scientist Jiang Xiaowei becoming a new member – the first scholar from the cloud computing field and Alibaba's second scientist to be honored.

HPCA, alongside ISCA and MICRO, is one of the three top international conferences on computer architecture, having driven many system technologies over the past 26 years, such as SMT in the late 1990s, virtualization in 2005, and non‑volatile memory in 2010. Inductees are scholars who have made outstanding contributions to computing, and only five Chinese scholars have previously received this honor, including Jiang’s colleague, senior researcher Xie Yuan.

Jiang earned his Ph.D. in Computer Engineering from North Carolina State University in 2009 and joined Alibaba Cloud’s server R&D team in 2014, focusing on CPU architecture, chip design, networking, and virtualization. After joining, he led China’s first custom X86 CPU project, making Alibaba the nation’s first technology company to develop a custom CPU.

In a recent HPCA‑published paper, Jiang systematically presented Alibaba Cloud’s self‑developed high‑performance AI server cluster EFLOPS, which can accelerate training of million‑class classification models by four times and support ten‑million‑class models; it also improves translation model accuracy while cutting training time from 100 hours to 12 hours. The cluster is now offered through Alibaba Cloud’s PAI deep‑learning platform. Additionally, Jiang’s team identified six bottlenecks for graph computing on CPU architectures and proposed optimization strategies, laying theoretical foundations for future chip and server designs.

Alibaba Cloud’s global influence continues to grow, serving users in over 200 countries and regions, providing cloud computing, AI, and IoT services to more than half of China’s listed companies and 80 % of its tech enterprises. Last year, Alibaba’s core systems were 100 % cloud‑based, handling Double‑11 traffic peaks, with AI calls exceeding 10 trillion per day and serving a billion users worldwide.

Talent underpins Alibaba’s technical depth: 60 000 scientists and engineers form the company’s technical foundation, with one‑third of partners having a technical background and 60 % of its 100 000 employees being technical staff. Numerous scientists have earned prestigious honors such as IEEE Fellow, academy memberships, and distinguished scientist awards.

Since 2009, Alibaba has persisted in developing its own Feitian cloud operating system, achieving a decade‑long breakthrough that propelled Chinese cloud computing from zero to one. In recent years, the company has entered a “hard‑core” technology explosion phase, covering chips, servers, quantum computing, blockchain, IoT, and more, establishing the broadest and deepest technical portfolio among Chinese tech firms.

High Performance ComputingCPU designAlibaba CloudHPCAAI ServersHall of Fame
Alibaba Cloud Infrastructure
Written by

Alibaba Cloud Infrastructure

For uninterrupted computing services

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.