ChinAI #179: The East-West Compute Transfer Project (东数西算)
China's latest mega infrastructure project
Greetings from a world where…
Ideals are plump, reality is skinny (理想很丰满，现实很骨感)
…As always, the searchable archive of all past issues is here. Please please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay support access for all AND compensation for awesome ChinAI contributors).
Feature Translation: East-West Compute Transfer project — what are the obstacles?
Context: In February 2022, China’s National Development and Reform Commission announced plans to construct 10 data center clusters. According to People’s Daily (English), the “The move signals that the country has officially started to implement its east-to-west computing resource transfer project for better development of digital economy.”
What is the “East-West Compute Transfer Project?” From ChinAI #159:
The motivation behind the project: Most of China's Internet and big data companies are located in eastern, coastal regions (Beijing, Tianjin, Shanghai, Zhejiang, Jiangsu, etc). Naturally, that's where they base their data centers too. But electricity costs are very high in these areas. Western provinces, such as Inner Mongolia, Gansu, and Guizhou, on the other hand, have much lower electricity costs, so the idea is to construct more data centers in these areas.
This is not a new idea. As early as 2013, Amazon AWS deployed a data center in Ningxia (autonomous region in northwestern China) while operating their service portal in Beijing. Yet this model has rarely found success. Why, and how will this major infrastructure project affect things? This week’s article (link to original Chinese) comes from Zhuzi talks carbon neutrality (朱子说碳中和), a research team that has been following data center and energy issues for more than 20 years. It was first published in June 2021.
Back when the author worked for an Internet giant in 2008 and was tasked with investigating site selection for data centers, here was the original vision:
“We were like science students with romantic ideas: base a site in Gansu, use plentiful solar and wind energy to power the data center, and let the melting snow naturally cool the surplus heat from data centers.”
Note the intersection of three strategic priorities: improve economic efficiency, reduce carbon emissions, and reduce energy dependence.
The author recounts obstacles his team faced:
Transmitting info from western data centers to eastern front-end services increases latency and delays
Companies liked keeping data centers in Beijing, Shanghai, Guangzhou, and Shenzhen because they could immediately deal with issues that came up; a common sentiment: “western regions are so far away, it’s hard to deal with.”
Customers just weren’t willing to buy cloud services if they were deployed in western data centers: “even if western data centers gave a lower price, 99% of customers still do not pay.”
What’s changed since then?
The author argues that technology, capital, and resources no longer hold back east-west compute transfer; now, it’s just “an ideological issue of decision makers on the application side” (the third obstacle above)
In this article’s view, the Chinese system is set up to resolve these types of issues. In September 2020, the government’s announcement of major goals to reduce carbon emissions and reach carbon neutrality before 2060 added momentum to the East-West Compute Transfer project.
Jeff’s comment: what hasn’t changed is the need for close location of data centers and cloud service users for real-time computations like providing search results or other AI-enabled services. It’ll be interesting to see whether companies can strike that balance, with western data centers doing tasks with less stringent latency requirements like background processing, offline analysis, and system backup.
FULL TRANSLATION: East-West Compute Transfer project: What are the obstacles
ChinAI Links (Four to Forward)
From the blurb by longreads for this High Country News article:
Aimee Towi Mae Tang, a fourth-generation Chinese New Mexican, felt disconnected from her Chinese roots. Amid a rise of anti-Asian hate crimes across the U.S., she wanted a better understanding of her own identity, which included learning how her family had settled in Albuquerque. Born in China and new to Albuquerque himself, journalist Wufei Yu decides to help Tang learn more about her family’s history, and in doing so, perhaps find his own place in a new city. Yu visits the National Archives in the San Francisco Bay Area to dig through documents: “For two days, those 400-plus colorful pages became my world — passenger arrival lists, immigration records, business filings and legal case files, dotted with Chinese characters”… Yu pieces together the story of Tang’s great-grandfather…a young man from South China who made the journey to San Francisco and fought to stay in America during its anti-Chinese immigration crackdown. He went on to become a businessman in Albuquerque, owning for a time one of the best grocery stores in town until its Chinese community was pushed out. “If my great-grandpa were allowed to have land,” Tang says in the piece, “the Tang family and Chinese Americans could have owned downtown Albuquerque.”
I gave some comments in this article by Eduardo Baptista on “one person, one file” software, based on a Reuters review of 50 tenders by Chinese police units and Party bodies trying to upgrade their surveillance networks. Some details follow:
Nine of the tenders indicated the software would be used with facial recognition technology that could, the documents specified, identify whether a passerby was Uyghur, connecting to early warning systems for the police and creating archives of Uyghur faces.
One tender published in February 2020 by a Party organ responsible for an area in the southeastern island province of Hainan, for instance, sought a database of Uyghur and Tibetan residents to facilitate “finding the information of persons involved in terrorism.”
Should-read: The grassroots intellectuals
In Dissent, Sebastian Veg asks whether there’s any point in paying attention to Chinese intellectuals. The conclusion: “China’s social and intellectual spheres remain less monolithic than the tightly controlled public transcripts would suggest, and their possibilities deserve our continued attention.”
Rui Zhong, a program associate at the Wilson Center, interviews Shazeda Ahmed, PhD candidate at Berkeley, on how Chinese courtrooms are utilizing AI, Chinese companies’ reactions to algorithm discrimination, and framing U.S.-China technological competition.
Thank you for reading and engaging.
These are Jeff Ding's (sometimes) weekly translations of Chinese-language musings on AI and related topics. Jeff is a postdoctoral fellow at Stanford's Center for International Security and Cooperation, sponsored by Stanford's Institute for Human-Centered Artificial Intelligence.
Check out the archive of all past issues here & please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay for a subscription will support access for all).
Any suggestions or feedback? Let me know at email@example.com or on Twitter at @jjding99