image
By Asia Education Review Team , Friday, 20 September 2024

Alibaba Cloud Empowers Customers with Large Models and Robust Infrastructure

  • Alibaba Cloud, the digital technology and intelligence backbone of Alibaba Group, announced at its annual Apsara Conference that its proprietary foundational model, Tongyi, has gained over 300,000 customers through its generative AI platform, Model Studio. This marks a substantial increase from the 90,000 customers reported in May.

    By utilizing Alibaba Cloud's Tongyi large models and its scalable global cloud infrastructure, businesses in the automotive, gaming, travel, and IT sectors have improved customer experiences and created innovative solutions. Additionally, Alibaba Cloud announced a partnership with the International Centre for Higher Education Innovation under UNESCO (UNESCO-ICHEI) to extend the advantages of generative AI to higher education institutions through digital training initiatives.

    "Our collaborations with industry-leading organizations demonstrate our dedication to delivering enhanced customer experiences and boosting enterprise efficiency through innovative solutions. Together with our partners, we want to empower more businesses and individuals to unlock the potential of generative AI through digital training and accessible cloud solutions", said Jingren Zhou, CTO of Alibaba Cloud.

    Alibaba Cloud today unveiled its large multimodal model (LMM) solution designed for automotive applications and co-developed with NVIDIA and Banma, Alibaba's intelligent cockpit solution provider. This solution, which will be made available to China-based automakers, aims to deliver a more interactive experience for car owners and redefine the smart mobility journey.

    Through this collaboration, Alibaba Cloud's suite of proprietary large language models, including the Qwen2-7B language model and the Qwen2-VL vision language model, has been seamlessly integrated with the NVIDIA DRIVE AGX Orin platform for autonomous vehicles. Furthermore, NVIDIA's model acceleration technology has greatly reduced computational costs and minimized latency in the real-time processing of complex tasks by Alibaba Cloud models. This ensures a smooth and uninterrupted intelligent experience for both drivers and passengers.

    This initiative marks the first integration of Alibaba's large models into NVIDIA's automotive platform. Leveraging Qwen's advanced capabilities in handling complex inquiries and processing visual intelligence, in-car voice assistants will be able to engage in dynamic, multi-turn conversations and even offer recommendations, ranging from providing information about nearby landmarks to proactively suggesting car headlights be turned on during snowy conditions.

    As part of the LMM solution, Alibaba Cloud's Mobile Agent a multimodal AI agent designed for edge devices will enhance the capabilities of in-car voice assistants. Leveraging the Mobile Agent’s strengths in complex task planning, self-reflection, and decision-making, users will be able to easily execute voice commands, such as ordering milkshakes through a food delivery app, leading to more enriching and enjoyable in-car experiences. Looking ahead, Alibaba Cloud plans to collaborate with NVIDIA to adapt Qwen models for NVIDIA DRIVE Thor, the next-generation centralized car compute platform that integrates advanced driver assistance, autonomous driving, and AI cockpit capabilities into a single safe and secure system.