As the AI revolution continues to accelerate, NVIDIA founder and CEO Jensen Huang made a powerful appearance at GTC China 2017, showcasing cutting-edge technology aimed at scaling artificial intelligence adoption across industries. The event attracted over 3,500 scientists, engineers, and media professionals, marking a significant milestone in NVIDIA’s global tour.
In his keynote speech, Huang highlighted how major internet companies are integrating deep learning into services used by billions of people, emphasizing the exponential growth of AI inference workloads. For instance, daily inference calculations in video processing reach 300 billion, while speech recognition processes over 50 billion, and language translation handles more than 700 billion operations per day.
To meet this demand, NVIDIA introduced TensorRT 3, the world's first programmable AI inference software. Designed for production environments, it significantly boosts performance from the cloud to edge devices like self-driving cars and robots, while also reducing costs. When paired with NVIDIA GPUs, TensorRT 3 enables ultra-fast and efficient AI processing across various frameworks, including image and speech recognition, natural language processing, and personalized recommendations.
According to Huang, the combination of TensorRT 3 and Tesla V100 GPU can process up to 5,700 images per second, compared to just 140 images per second using traditional CPUs. A single server equipped with the latest NVIDIA GPU outperforms 150 CPU servers, effectively replacing the computing power of 13 data centers powered by CPUs alone.
In addition to TensorRT 3, NVIDIA also launched other AI acceleration tools, such as the DeepStream SDK for real-time video analytics and CUDA 9, which enhances high-performance computing and deep learning applications.
Huang announced that leading Chinese cloud providers like Alibaba, Baidu, and Tencent have deployed Tesla V100 GPU accelerators in their services. Major OEMs, including Huawei, Inspur, and Lenovo, have adopted NVIDIA’s HGX server architecture, using Tesla V100 GPUs to build next-generation accelerated data centers.
NVIDIA’s DRIVE platform, an open computing solution for autonomous vehicles, has been adopted by 145 startups. DRIVE AV integrates data from cameras, radar, and lidar, using deep learning and computer vision algorithms to achieve L4 and L5 autonomy. Partners can leverage any part of the platform to develop their own solutions.
NVIDIA also unveiled Xavier, the world’s first self-contained AI processor, designed for autonomous systems. It will be available to early partners in Q1 2018 and fully launched in Q4 2018. Xavier will serve as the next-generation Jetson system-on-chip. JD.com has already adopted the Jetson platform to enhance its AI-driven logistics and delivery systems.
JD.com is developing JDrone and JDrover—unmanned delivery tools powered by NVIDIA Jetson and TensorRT. With over 30 billion packages delivered annually, JD plans to expand its fleet to one million drones by 2022.
NVIDIA also announced new partnerships under its Metropolis initiative, including Hikvision, the world’s top intelligent video solutions provider. Alongside Alibaba and Huawei, these companies are building smarter cities using NVIDIA’s AI-powered video analytics. By 2020, over one billion city cameras will be leveraged for traffic management, emergency alerts, and missing person searches.
Looking ahead, NVIDIA will continue sharing its AI vision with developers in Munich, Tel Aviv, Taipei, Washington, and Tokyo.
Low Speed and Middle Speed Generators with Low speed engine or middle speed engine
. Low speed diesel generators, 500rpm, 600rpm
. Middle speed diesel generators, 720rpm, 750rpm, 900rpm, 1000rpm, 1200rpm
. Voltage: 3kV, 3.3kV, 6kV, 6.3kV, 6.6kV, 10kV, 10.5kV, 11kV,13.8kV
. Frequency: 50Hz, 60Hz
Low Speed Generator,Middle Speed Generator,1000Rpm Diesel Generator,500Rpm Diesel Generator
Guangdong Superwatt Power Equipment Co., Ltd , https://www.swtgenset.com