عربي English Türkçe اردو فارسى
Logo
Latest News
Technology

ByteDance Plans New AI Model Trained With Huawei Chips

The ByteDance logo is seen at the company's office in Shanghai, China, July 4, 2023. REUTERS/Aly Song/File Photo
The ByteDance logo is seen at the company's office in Shanghai, China, July 4, 2023. REUTERS/Aly Song/File Photo
TT
TT

ByteDance Plans New AI Model Trained With Huawei Chips

The ByteDance logo is seen at the company's office in Shanghai, China, July 4, 2023. REUTERS/Aly Song/File Photo
The ByteDance logo is seen at the company's office in Shanghai, China, July 4, 2023. REUTERS/Aly Song/File Photo

TikTok's Chinese parent ByteDance plans to develop an AI model trained primarily with chips from compatriot Huawei Technologies, said three people familiar with the matter, as US curbs turn the social media giant homeward in search of chips.
ByteDance has diversified to domestic suppliers of chips used in artificial intelligence and accelerated development of its own since the US in 2022 started restricting exports of advanced AI chips such as from market leader Nvidia, Reuters said.
AI has become central to the technology industry with firms in sectors as varied as gaming and e-commerce differentiating offerings through the integration of custom AI models - programs that employ pattern recognition to make decisions.
ByteDance's next step in the AI race is to use Huawei's Ascend 910B chip to train a large-language AI model, said the people, declining to be identified as the plan is confidential.
A fourth person also said ByteDance is planning a new AI model but could not say whether it will use Huawei chips.
ByteDance already uses the Ascend 910B primarily for less computationally intensive inference tasks, which involve pre-trained AI models making predictions, the three people and a separate source said.
Training AI models is far more demanding and requires huge amounts of data, necessitating the use of ultra-high-performance chips such as Nvidia's premium graphics processing units.
The new model's capability and complexity, measured by its computing parameters, will be less powerful than ByteDance's existing AI model Doubao, one of the people said.
ByteDance did not reply to a request for comment. Michael Hughes, a TikTok spokesman in Washington D.C., said, speaking on behalf of ByteDance, "The entire premise here is wrong. No new model is being developed".
Huawei did not reply to Reuters' requests for comment.
TIGHT SUPPLY
ByteDance has ordered more than 100,000 Ascend 910B chips this year but has received fewer than 30,000 as of July, a pace too slow to meet company needs, one of the people said.
The constrained supply and limited computing power versus Nvidia's China-available chips have prevented ByteDance from setting a timeline for the new model, two of the people said.
ByteDance's current AI technology is used in its flagship large-language model launched in August 2023 and rebranded as chatbot Doubao, and in many other applications including a text-to-video tool Jimeng. It introduced two video-focused Doubao models this month to compete with OpenAI.
Use of such applications has ballooned since early this year, with ByteDance's chatbot becoming one of China's most popular apps with more than 10 million monthly active users.
The increased emphasis on AI has made ByteDance one of the largest buyers of Huawei's AI chips, the three people said.
It is also the biggest buyer of Nvidia's H20 AI chip, which the U.S. chipmaker tailored for the China market in response to trade restrictions, said two of the people. The TikTok owner is also Microsoft's biggest client in Asia for Nvidia chips accessible via cloud computing, said two separate sources.
Reuters previously reported that ByteDance allocated $2 billion for Nvidia chips last year.
Nvidia declined to comment. Microsoft did not reply to a request for comment.

Most Viewed

 
 
Technology

Lenovo Unveils AI Agent to Bridge PCs, Phones and Wearables at CES

Attendees gather for the annual Consumer Electronics Show (CES) in Las Vegas, Nevada, on January 6, 2026. (AFP)
Attendees gather for the annual Consumer Electronics Show (CES) in Las Vegas, Nevada, on January 6, 2026. (AFP)
TT
TT

Lenovo Unveils AI Agent to Bridge PCs, Phones and Wearables at CES

Attendees gather for the annual Consumer Electronics Show (CES) in Las Vegas, Nevada, on January 6, 2026. (AFP)
Attendees gather for the annual Consumer Electronics Show (CES) in Las Vegas, Nevada, on January 6, 2026. (AFP)

Lenovo, the world's top PC maker, unveiled its own AI assistant Tuesday at the CES tech show in Las Vegas, promising a tool that follows users seamlessly across laptops, smartphones and connected devices.

The Beijing-based company commanded 28 percent of global PC market share in the third quarter of 2025, ahead of rivals HP at 21.5 percent and Dell at 14.5 percent, according to US research firm Gartner.

Lenovo's new artificial intelligence agent, dubbed Qira, is designed as an autonomous interface capable of performing tasks rather than simply generating content on demand, a move Lenovo hopes will showcase the breadth of its product portfolio.

Unlike rivals focused on single categories, Lenovo was the only major manufacturer whose offering spanned laptops, tablets and smartphones -- under its Motorola brand, acquired in 2014 -- as well as servers and even supercomputers.

The company also unveiled prototypes of connected glasses and an AI-powered pendant, still in testing, that captures "important moments" with user consent by recording conversations, said Motorola's Angelina Gomez.

Codenamed the AI Perceptive Companion, the pendant features a microphone and camera and "sees what you see and hears what your hear," Lenovo vice president Luca Rossi told reporters.

An interaction with Qira can start via the pendant, continue on a smartphone and end on a laptop, with the agent retaining user context across devices.

It can summarize the highlights of a user's day, draft and send emails, or even select photos from archives to post on social media.

Lenovo stressed it is not positioning Qira as a rival to Microsoft's Copilot and announced the integration of Copilot into Motorola smartphones.

For major hardware makers, the challenge now is proving the utility of generative AI in everyday applications rather than simply flaunting cutting-edge tech.

Amid lingering geopolitical tensions with Washington, Lenovo was the only Chinese firm to take center stage at CES, choosing Las Vegas's futuristic Sphere venue for its showcase.

Executives emphasized the company's global footprint, with most revenue generated outside China and several top managers from overseas.

Technology

At CES, Auto and Tech Companies Transform Cars into Proactive Companions

The Afeela Prototype 2026 displayed during a Sony Honda Mobility news conference ahead of the CES tech show, Monday, Jan. 5, 2026, in Las Vegas. (AP)
The Afeela Prototype 2026 displayed during a Sony Honda Mobility news conference ahead of the CES tech show, Monday, Jan. 5, 2026, in Las Vegas. (AP)
TT
TT

At CES, Auto and Tech Companies Transform Cars into Proactive Companions

The Afeela Prototype 2026 displayed during a Sony Honda Mobility news conference ahead of the CES tech show, Monday, Jan. 5, 2026, in Las Vegas. (AP)
The Afeela Prototype 2026 displayed during a Sony Honda Mobility news conference ahead of the CES tech show, Monday, Jan. 5, 2026, in Las Vegas. (AP)

In a vision of the near future shared at CES, a girl slides into the back seat of her parents' car and the cabin instantly comes alive. The vehicle recognizes her, knows it’s her birthday and cues up her favorite song without a word spoken.

“Think of the car as having a soul and being an extension of your family,” Sri Subramanian, Nvidia's global head of generative AI for automotive, said Tuesday.

Subramanian's example, shared with a CES audience on the show's opening day in Las Vegas, illustrates the growing sophistication of AI-powered in-cabin systems and the expanding scope of personal data that smart vehicles may collect, retain and use to shape the driving experience.

Across the show floor, the car emerged less as a machine and more as a companion as automakers and tech companies showcased vehicles that can adapt to drivers and passengers in real time — from tracking heart rates and emotions to alerting if a baby or young child is accidentally left in the car.

Bosch debuted its new AI vehicle extension that aims to turn the cabin into a “proactive companion.” Nvidia, the poster child of the AI boom, announced Alpamayo, its new vehicle AI initiative designed to help autonomous cars think through complex driving decisions. CEO Jensen Huang called it a “ChatGPT moment for physical AI.”

But experts say the push toward a more personalized driving experience is intensifying questions about how much driver data is being collected.

“The magic of AI should not just mean all privacy and security protections are off,” said Justin Brookman, director of marketplace policy at Consumer Reports.

Unlike smartphones or online platforms, cars have only recently become major repositories of personal data, Brookman said. As a result, the industry is still trying to establish the “rules of the road” for what automakers and tech companies are allowed to do with driver data.

That uncertainty is compounded by the uniquely personal nature of cars, Brookman said. Many people see their vehicles as an extension of themselves — or even their homes — which he said can make the presence of cameras, microphones and other monitoring tools feel especially invasive.

“Sometimes privacy issues are difficult for folks to internalize,” he said. “People generally feel they wish they had more privacy but also don’t necessarily know what they can do to address it.”

At the same time, Brookman said, many of these technologies offer real safety benefits for drivers and can be good for the consumer.

On the CES show floor, some of those conveniences were on display at automotive supplier Gentex’s booth, where attendees sat in a mock six-seater van in front of large screens demonstrating how closely the company’s AI-equipped sensors and cameras could monitor a driver and passengers.

“Are they sleepy? Are they drowsy? Are they not seated properly? Are they eating, talking on phones? Are they angry? You name it, we can figure out how to detect that in the cabin,” said Brian Brackenbury, director of product line management at Gentex.

Brackenbury said it's ultimately up to the car manufacturers to decide how the vehicle reacts to the data that's collected, which he said is stored in the car and deleted after the video frames, for example, have been processed. "

“One of the mantras we have at Gentex is we're not going to do it just because we can, just because the technology allows it,” Brackebury said, adding that “data privacy is really important.”

Technology

Nvidia CEO Huang Says Next Generation of Chips Is in Full Production

Nvidia CEO Jensen Huang speaks during a Nvidia keynote address at CES 2026, an annual consumer electronics trade show, in Las Vegas, Nevada, US January 5, 2026. (Reuters)
Nvidia CEO Jensen Huang speaks during a Nvidia keynote address at CES 2026, an annual consumer electronics trade show, in Las Vegas, Nevada, US January 5, 2026. (Reuters)
TT
TT

Nvidia CEO Huang Says Next Generation of Chips Is in Full Production

Nvidia CEO Jensen Huang speaks during a Nvidia keynote address at CES 2026, an annual consumer electronics trade show, in Las Vegas, Nevada, US January 5, 2026. (Reuters)
Nvidia CEO Jensen Huang speaks during a Nvidia keynote address at CES 2026, an annual consumer electronics trade show, in Las Vegas, Nevada, US January 5, 2026. (Reuters)

Nvidia CEO Jensen Huang said on Monday that the company’s next generation of chips is in "full production," saying they can deliver five times the artificial-intelligence computing of the company’s previous chips when serving up chatbots and other AI apps.

In a speech at the Consumer Electronics Show in Las Vegas, the leader of the world's most valuable company revealed new details about its chips, which will arrive later this year and which Nvidia executives told Reuters are already in the company's labs being tested by AI firms, as Nvidia faces increasing competition from rivals as well as its own customers.

The Vera Rubin platform, made up of six separate Nvidia chips, is expected to debut later this year, with the flagship server containing 72 of the company’s graphics units and 36 of its new central processors.

Huang showed ‌how they can ‌be strung together into "pods" with more than 1,000 Rubin chips and said they could improve ‌the ⁠efficiency of generating ‌what are known as "tokens" - the fundamental unit of AI systems - by 10 times.

To get the new performance results, however, Huang said the Rubin chips use a proprietary kind of data that the company hopes the wider industry will adopt.

"This is how we were able to deliver such a gigantic step up in performance, even though we only have 1.6 times the number of transistors," Huang said.

While Nvidia still dominates the market for training AI models, it faces far more competition - from traditional rivals such as Advanced Micro Devices as well as customers like Alphabet's Google - in delivering the fruits of those models to hundreds of millions of users of chatbots and other technologies.

Much of ⁠Huang’s speech focused on how well the new chips would work for that task, including adding a new layer of storage technology called "context memory storage" aimed at helping chatbots ‌provide snappier responses to long questions and conversations.

Nvidia also touted a new generation ‍of networking switches with a new kind of connection called co-packaged ‍optics. The technology, which is key to linking together thousands of machines into one, competes with offerings from Broadcom and Cisco ‍Systems.

Nvidia said that CoreWeave will be among the first to have the new Vera Rubin systems and that it expects Microsoft, Oracle, Amazon and Alphabet to adopt them as well.

In other announcements, Huang highlighted new software that can help self-driving cars make decisions about which path to take - and leave a paper trail for engineers to use afterward. Nvidia showed research about software, called Alpamayo, late last year, with Huang saying on Monday it would be released more widely, along with the data used to train it so that automakers can make evaluations.

"Not only do we open-source the models, we also open-source the data that we use ⁠to train those models, because only in that way can you truly trust how the models came to be," Huang said from a stage in Las Vegas.

Last month, Nvidia scooped up talent and chip technology from startup Groq, including executives who were instrumental in helping Alphabet's Google design its own AI chips. While Google is a major Nvidia customer, its own chips have emerged as one of Nvidia's biggest threats as Google works closely with Meta Platforms and others to chip away at Nvidia's AI stronghold.

During a question-and-answer session with financial analysts after his speech, Huang said the Groq deal "won't affect our core business" but could result in new products that expand its lineup.

At the same time, Nvidia is eager to show that its latest products can outperform older chips like the H200, which US President Donald Trump has allowed to flow to China. Reuters has reported that the chip, which was the predecessor to Nvidia's current "Blackwell" chip, is in high demand in China, which has alarmed China hawks across the US political spectrum.

Huang told financial analysts after his keynote ‌that demand is strong for the H200 chips in China, and Chief Financial Officer Colette Kress said Nvidia has applied for licenses to ship the chips to China but was waiting for approvals from the US and other governments to ship them.