Kneron launches latest AI chip KL730 to drive large-scale application of lightweight GPT solutions

The progress in energy efficiency of KL730 solves the biggest bottleneck in the implementation of artificial intelligence models - energy costs, which are 3 to 4 times higher than those in the industry and previous Nerner chips

The KL730 chip supports the most advanced lightweight GPT large language model, such as nanoGPT, and provides an effective computing power of 0.35 - 4 tera per second

AI company Kneron today announced the release of the KL730 chip, which integrates car-grade NPU and image signal processing (ISP), empowering safe and low-energy AI capabilities to edge servers, smart homes and automotive assisted driving systems. and other various application scenarios. San Diego-based Kneron is best known for its groundbreaking neural processing units (NPUs)

Kneron’s latest chip KL730 is designed to realize artificial intelligence functions and has made breakthroughs in a number of energy-saving and security technologies. The chip has a multi-channel interface that can seamlessly access a variety of digital signals, such as images, videos, audio and millimeter waves, to support the development of artificial intelligence applications in various industries

The chip also solves one of the current widespread bottlenecks in artificial intelligence: the high system cost caused by widespread inefficient hardware.

KL730 has made a huge breakthrough in the research and development of energy efficiency. Compared with previous Kneron chips, its energy efficiency has increased by 3 to 4 times, and is 150%~200% higher than the main products in the same industry

Kneron founder and CEO Liu Juncheng said that KL730 will become an innovator in edge AI, providing powerful AI capabilities to all walks of life through its unprecedented efficiency and support for frameworks such as Transformer, while ensuring data security and privacy. Protect and unleash the full potential of artificial intelligence

Kneron focuses on edge AI and has successfully developed a series of lightweight and scalable AI chips to safely promote the development of AI capabilities. In 2021, Kneron released the KL530, the first edge AI chip to support the Transformer neural network architecture. The Transformer neural network architecture is the basis of all GPT models. The KL730 chip further enriches the product series, providing effective computing power of 0.35 - 4 tera per second, and expanding the ability to support the most advanced lightweight GPT large language models (such as nanoGPT)

KL730 is a uniquely positioned chip that can improve security in the AIot field and enable users to run GPT models partially or completely offline on terminal devices. Paired with Kneo, Kneron’s private secure edge AI network, this chip enables AI to run on users’ edge devices, thereby better protecting data privacy. These applications are widely used in various industries, including enterprise server solutions, smart driving vehicles, and AI-powered medical equipment. Enhanced security enables devices to work better together and keeps data safe. For example, engineers can design new semiconductor chips without sharing sensitive data with data centers run by large cloud companies

Since its establishment in 2015, Kneron has won wide recognition in the industry for its reconfigurable NPU architecture and has won multiple awards, including the IEEE Darlington Award. Kneron chips have been successfully used in terminal products in multiple industries, covering fields such as AIoT, smart driving and edge servers. Partners include Toyota, Quanta Electronics, Chunghwa Telecom, Panasonic, Hanwha and many other well-known companies

The KL730 will be sampling soon to device manufacturers, learn more and explore the unlimited potential of the KL730

About persistent performance

Founded in 2015 and headquartered in San Diego, USA, Kneron is the world's leading manufacturer of full-scale edge AI computing solutions. Through its self-developed efficient and lightweight reconfigurable neural network architecture, Kneron has successfully solved the three major problems faced by edge AI devices, including latency, security and cost, thereby realizing ubiquitous AI. Up to now, Kneron has received more than 140 million US dollars in financing, and has received investment from Horizons Investment, Sequoia Capital, Qualcomm, Hon Hai, Lite-On Technology, Winbond Electronics, Macronix Electronics, ADATA Technology, Quanke Technology Waiting for support from multiple investors

The above is the detailed content of Kneron launches latest AI chip KL730 to drive large-scale application of lightweight GPT solutions. For more information, please follow other related articles on the PHP Chinese website!