NVIDIA releases its latest automotive chip, how powerful is it?

Publisher:程序界的行者Latest update time:2021-04-13 Keywords:NVIDIA Reading articles on mobile phones Scan QR code
Read articles on your mobile phone anytime, anywhere

NVIDIA, a processor technology giant, recently announced its next-generation artificial intelligence (AI) processor for self-driving cars, NVIDIA DRIVE Atlan, at its GTC annual conference. This processor can provide more than 1,000 TOPS of computing power, targeting the models that major car manufacturers will launch in 2025. It combines AI and BlueField technology on a single chip. The new SoC will provide self-driving cars with more than 1,000 TOPS of computing power and data center-level security.



Processor technology giant NVIDIA recently announced its next-generation artificial intelligence (AI) processor NVIDIA DRIVE Atlan for self-driving cars at its annual GTC conference. This processor can provide more than 1,000 TOPS of computing power and is targeted at models that major automakers will launch in 2025.


NVIDIA pointed out that NVIDIA DRIVE Atlan on a single chip is the latest product in NVIDIA's centralized computing blueprint for self-driving cars, combining AI and software with the latest computing, networking, and security to achieve unprecedented performance and safety.


DRIVE Atlan will use NVIDIA's next-generation GPU architecture, new Arm CPU cores, and deep learning and computer vision accelerators. With data center-level performance, Atlan will provide automakers with sufficient computing power to build software-defined vehicles with rich programmable capabilities and the ability to be permanently upgraded through secure over-the-air (OTA) updates.



NVIDIA DRIVE Orin is the world’s highest-performance, most advanced processor for autonomous vehicles and robotics, delivering up to 254 TOPS of computing power.


“The transportation industry needs a reliable computing platform that can last for decades. Software investments are too large to be repeated for every vehicle,” said Jensen Huang, founder and CEO of NVIDIA. “NVIDIA DRIVE is the most advanced AI and self-driving computing platform with a rich global software and developer ecosystem and architectural compatibility from generation to generation. Today we’re announcing a new addition to our product roadmap. The new DRIVE Atlan is truly a technological marvel that brings together NVIDIA’s strengths in AI, automotive, robotics, safety and the BlueField secure data center to enable our customers to build a safe fleet of self-driving cars.”


DRIVE Atlan will integrate the NVIDIA BlueField data processing unit (DPU), which provides a wide range of advanced networking, storage and security services to support complex computing and AI workloads in self-driving cars. BlueField provides single-chip programmability of a complete data center infrastructure, and also adds a secure isolation zone to prevent data leakage and cyberattacks. NVIDIA designed the DRIVE Atlan chip from scratch to handle the large number of AI applications running simultaneously in self-driving cars in a safe and reliable way.


NVIDIA said that the NVIDIA DRIVE Xavier (30 TOPS) processor previously used for self-driving cars is now used in mass-produced cars and trucks, and major automakers have already chosen to use NVIDIA DRIVE Orin (254 TOPS) in their production lines starting in 2022. NVIDIA DRIVE Atlan targets cars produced in 2025 and later, and will expand the leadership of the NVIDIA DRIVE system-on-chip product line.


Since the advent of NVIDIA DRIVE Atlan, Orin and Xavier have been programmable through open source CUDA and TensorRT APIs and libraries, allowing developers to leverage their investment across multiple generations as they build the roadmap for future production self-driving cars.


Keywords:NVIDIA Reference address:NVIDIA releases its latest automotive chip, how powerful is it?

Previous article:Introduction to automotive semiconductor safety regulations and packaging quality inspection (Part 2)
Next article:Velodyne and Ansys collaborate on next-generation automotive lidar sensors

Recommended ReadingLatest update time:2024-11-16 11:47

Nvidia releases Thor, a car chip
At 11:30 pm on September 20, NVIDIA's Fall GTC kicked off on time. NVIDIA CEO Jensen Huang delivered a speech live online, announcing: "Atlan is no longer number one... it will be replaced by Thor." This SoC chip has 77 billion transistors inside, which can achieve 2000TOPS of AI computing power, or 2000TFLOPs. The
[Automotive Electronics]
Nvidia releases Thor, a car chip
Interpreting Tesla HW4.0: Which is better, the second-generation FSD or Nvidia Orin?
@greentheonly, who goes by the handle “Green God” on Twitter, recently released photos of the actual hardware of Tesla’s latest HW4.0. This information is very valuable, and we can get a glimpse of the latest progress and related technical level of the second-generation FSD chip in HW4.0. The h
[Automotive Electronics]
Nvidia acquires data storage and management platform SwiftStack to accelerate edge deployment
According to foreign media reports, Nvidia announced today that it has acquired the data storage and management platform SwiftStack.                                                  SwiftStack is a software-centric data storage and management platform that supports public cloud, on-premises, and edge deployments.   S
[Internet of Things]
Nvidia acquires data storage and management platform SwiftStack to accelerate edge deployment
Nvidia and universities work together to invest $70 million to build the world's fastest AI supercomputer
According to foreign media reports, NVIDIA has joined hands with the University of Florida in the United States to build the world's fastest higher education artificial intelligence (AI) supercomputer at the university.   NVIDIA defined the GPU and is the leader in computing with the invention of the GPU (graphics pro
[Embedded]
Nvidia and universities work together to invest $70 million to build the world's fastest AI supercomputer
Nvidia RTX 50 series graphics cards will be delayed! Need to re-tape out to improve yield
On September 3, according to media reports, the latest news from the supply chain showed that due to the need for re-tapeout (RTO) to improve the yield rate, the launch time of the RTX 50 series graphics cards will be delayed than originally planned, but there was no mention of when it will be released. NVIDIA'
[Home Electronics]
Microchip Accelerates Real-Time Edge AI Deployment with NVIDIA Holoscan Platform
PolarFire® FPGA Ethernet Sensor Bridge Provides Low-Power Multi-Sensor Bridging for NVIDIA Edge AI Platforms Function To help developers build artificial intelligence (AI)-driven sensor processing systems, Microchip Technology Inc. has released the PolarFire® FPGA-to-Ethernet Sensor Bridge supp
[Network Communication]
Microchip Accelerates Real-Time Edge AI Deployment with NVIDIA Holoscan Platform
Why did the US FTC reject Nvidia’s $75 billion antitrust settlement agreement for the acquisition of Arm?
Regulators usually block mergers when competing companies suddenly want to merge, but they have recently become more stringent about vertical mergers between companies that are not competing with each other.    A recent example is the Federal Trade Commission’s handling of chipmaker Nvidia’s controversial acquisition
[Semiconductor design/manufacturing]
ChatGPT is on fire for AI chips, NVIDIA: the demand is extremely large
“My phone calls are almost off the hook these days.” Talking about the recent popularity of ChatGPT, a senior executive from Nvidia China told a reporter from the Science and Technology Innovation Board Daily in a slightly complaining but undisguisedly happy tone. The logic is that AI chips, the core infrastructure th
[Internet of Things]
Latest Automotive Electronics Articles
Change More Related Popular Components

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号