Google Reportedly Weighs Large Data Center in Vietnam

FILE PHOTO: The logo for Google is seen at the Google Store Chelsea in Manhattan, New York City, US, November 17, 2021. REUTERS/Andrew Kelly/File Photo
FILE PHOTO: The logo for Google is seen at the Google Store Chelsea in Manhattan, New York City, US, November 17, 2021. REUTERS/Andrew Kelly/File Photo
TT

Google Reportedly Weighs Large Data Center in Vietnam

FILE PHOTO: The logo for Google is seen at the Google Store Chelsea in Manhattan, New York City, US, November 17, 2021. REUTERS/Andrew Kelly/File Photo
FILE PHOTO: The logo for Google is seen at the Google Store Chelsea in Manhattan, New York City, US, November 17, 2021. REUTERS/Andrew Kelly/File Photo

Alphabet's Google is considering building a large data center in Vietnam, a person briefed on the plans said, in what would be the first such investment by a big US technology company in the Southeast Asian nation.
Google is weighing setting up a "hyperscale" data center close to Ho Chi Minh City, Vietnam's southern economic hub, the source said, declining to be named because the information is not public.
The investment, the size of which the source did not specify, would be a shot in the arm for Vietnam which has so far failed to attract major overseas capital in data centers due to its patchy infrastructure, with large tech companies preferring to house their centers in rival nations in the region.
According to Reuters, it was not clear how quickly Google will reach a decision on an investment but the source said internal talks are on and the data center could be ready in 2027.
A spokesperson for Google declined to comment about the data center plan.
Hyperscale centers are the largest in the industry, with power consumption usually similar to that of a big city.
A hyperscale data center with power consumption capacity of 50 megawatts (MW) could cost between $300 million and $650 million, according to estimates based on data published by real estate consultant Jones Lang LaSalle in a report this year on data centers in Vietnam.
Google's move was motivated by the large number of its domestic and foreign cloud services clients in Vietnam and the country's expanding digital economy, the source said, noting the Southeast Asian nation was one of the fastest-growing markets for YouTube, Google's popular online video sharing platform.
Currently the top data center operators in Vietnam, based on computing space, are industrial investment firm IDC Becamex and telecommunications company VNPT, both Vietnamese state-owned enterprises, according to an internal market report by an industrial park in Vietnam seen by Reuters.
The Nikkei reported in May that Chinese e-commerce company Alibaba was considering building a data center in Vietnam. Alibaba did not reply to a request for comment.



Cerebras Launches AI Inference Tool to Challenge Nvidia

Cerebras Systems logo is seen in this illustration taken March 31, 2023. (Reuters)
Cerebras Systems logo is seen in this illustration taken March 31, 2023. (Reuters)
TT

Cerebras Launches AI Inference Tool to Challenge Nvidia

Cerebras Systems logo is seen in this illustration taken March 31, 2023. (Reuters)
Cerebras Systems logo is seen in this illustration taken March 31, 2023. (Reuters)

Cerebras Systems launched on Tuesday a tool for AI developers that allows them to access the startup's outsized chips to run applications, offering what it says is a much cheaper option than industry-standard Nvidia processors.

Access to Nvidia graphics processing units (GPUs) - often via a cloud computing provider - to train and deploy large artificial intelligence models used for applications such as OpenAI's ChatGPT can be difficult to obtain and expensive to run, a process developers refer to as inference.

"We're delivering performance that cannot be achieved by a GPU," Cerebras CEO Andrew Feldman told Reuters in an interview. "We're doing it at the highest accuracy, and we're offering it at the lowest price."

The inference portion of the AI market is expected to be fast-growing and attractive - ultimately worth tens of billions of dollars if consumers and businesses adopt AI tools.

The Sunnyvale, California-based company plans to offer several types of the inference product via a developer key and its cloud. The company will also sell its AI systems to customers who prefer to operate their own data centers.

Cerebras' chips - each the size of a dinner plate and called Wafer Scale Engines - avoid one of the issues with AI data crunching: the data crunched by large models that power AI applications typically won't fit on a single chip and can require hundreds or thousands of chips strung together.

That means Cerebras' chips can achieve speedier performances, Feldman said.

It plans to charge users as little as 10 cents per million tokens, which are one of the ways companies can measure the amount of output data from a large model.

Cerebras is aiming to go public and filed a confidential prospectus with the Securities and Exchange Commission this month, the company said.