Alibaba Group Holding's cloud computing services unit and semiconductor powerhouse Nvidia are collaborating on an artificial intelligence (AI) initiative that will enable Chinese automakers to advance the autonomous driving experience for smart vehicle owners.

Alibaba Cloud, the digital technology backbone of the e-commerce giant, on Friday announced that collaboration, as it unveiled a large multimodal model (LMM) solution for automotive applications that it co-developed with Nvidia and start-up Banma Network Technology, Alibaba's intelligent cockpit solution provider, at the Apsara Conference that runs until Saturday. Alibaba owns the South China Morning Post.

Hangzhou-based Alibaba Cloud's Qwen portfolio of proprietary large language models (LLMs) - including Qwen2-7B and Qwen2-VL - have been integrated with Nvidia's Drive AGX Orin platform for autonomous vehicles. LLMs are the technology underpinning generative AI services like ChatGPT.

Do you have questions about the biggest topics and trends from around the world? Get the answers with SCMP Knowledge, our new platform of curated content with explainers, FAQs, analyses and infographics brought to you by our award-winning team.

This marks the first integration of Alibaba Cloud's large AI models into Nvidia's Drive automotive platform, which major Chinese electric vehicle (EV) makers - including Li Auto, Great Wall Motors, Geely Auto's Zeekr Intelligent Technology and the EV unit of Xiaomi - are using to power their next-generation fleets.

Nvidia's Drive AGX Orin system-on-a-chip platform serves as the brain for intelligent automated-driving systems used by major Chinese electric vehicle makers. Photo: Handout alt=Nvidia's Drive AGX Orin system-on-a-chip platform serves as the brain for intelligent automated-driving systems used by major Chinese electric vehicle makers. Photo: Handout>

Nvidia's model acceleration technology has already significantly reduced computational costs and minimised latency in the real-time processing of complex tasks by Alibaba Cloud's AI models. This ensures a smooth and uninterrupted intelligent experience for both drivers and passengers, according to Alibaba Cloud.

"Together with our partners, we want to empower more businesses and individuals to unlock the potential of generative AI," Alibaba Cloud chief technology officer Zhou Jingren said at the event.

With Qwen's advanced capabilities in handling complex inquiries and processing visual intelligence, the new LMM solution will enable in-car voice assistants to engage in dynamic, multi-turn conversations, according to Alibaba Cloud. These assistants will also be able to offer recommendations, ranging from providing information about nearby landmarks to proactively suggesting car headlights be turned on during certain conditions.

As part of that LMM solution, Alibaba Cloud's Mobile Agent will enable vehicle owners to effortlessly execute voice commands, such as ordering milkshakes through a food delivery app, resulting in richer in-car experiences, according to Alibaba Cloud.

While stringent US export controls bar Nvidia from selling its most-advanced graphics processing units to mainland Chinese firms, the US company still has a thriving business on the mainland. In its financial year ended January 28, Nvidia said China was its third-largest market.

Alibaba Cloud's resources and capabilities are on display at the three-day Apsara Conference in Hangzhou, capital of eastern Zhejiang province, that runs until Saturday. Photo: Xinhua alt=Alibaba Cloud's resources and capabilities are on display at the three-day Apsara Conference in Hangzhou, capital of eastern Zhejiang province, that runs until Saturday. Photo: Xinhua>

Nvidia and Alibaba Cloud on Friday announced at Apsara that they plan to work on adapting Qwen LLMs on the US firm's Drive Thor, a next-generation centralised car compute platform that combines advanced driver assistance, autonomous driving and AI cockpit capabilities.

Santa Clara, California-based Nvidia was also working with Alibaba Cloud in terms of infrastructure "to meet people's extreme demand for computing power", Nvidia global vice-president He Tao said at a sub-forum hosted by the US company during the conference.

Nvidia and Alibaba Cloud are pursuing a number of LLM-related initiatives for accelerated computing, which use parallel processing to speed up work and increase energy efficiency across demanding applications, according to Nvidia.

In addition, the two companies are working to create customised solutions to help "traditional enterprises" move their LLM operations onto the cloud. Cloud computing technology enables enterprises to manage or distribute over the internet a range of software and other digital resources as an on-demand service, just like electricity from a power grid. These resources are stored inside data centres.

Alibaba Cloud on Friday said its proprietary foundational model Tongyi has attracted more than 300,000 customers via its generative AI platform Model Studio, a significant jump from 90,000 revealed in May.

This article originally appeared in the South China Morning Post (SCMP), the most authoritative voice reporting on China and Asia for more than a century. For more SCMP stories, please explore the SCMP app or visit the SCMP's Facebook and Twitter pages. Copyright © 2024 South China Morning Post Publishers Ltd. All rights reserved.

Copyright (c) 2024. South China Morning Post Publishers Ltd. All rights reserved.