Nvidia, Meet Everything About Autonomous Vehicles at 'GTC2017'
Published2017.05.02 10:55Presentations on the Impact of Artificial Intelligence on Autonomous Vehicles
Nvidia announced that it will prepare various autonomous vehicle-related sessions at 'GTC 2017 (GPU Technology Conference 2017)', the world's largest GPU developer conference, held in Silicon Valley from May 8th to 11th U.S. local time.
This year's GTC will feature conferences and hands-on sessions, exhibitions, demonstrations, and networking events. Among them, the autonomous vehicle software session, attended by numerous industry partners and experts, is expected to introduce how artificial intelligence is being utilized in autonomous vehicle software development.
First, Mercedes-Benz's head of Connected Car, User Interaction, and Telematics R&D will demonstrate how to implement artificial intelligence in vehicles equipped with embedded hardware for sensor processing and fusion solutions inside the vehicle.
Additionally, Roborace will introduce artificial intelligence technology related to Robocars and shed light on how software is defining the future of the automotive industry and racing sports.
Sessions showcasing autonomous vehicle implementation technology will also be held. Nvidia has arranged a session introducing NVIDIA DRIVE PX 2 and NVIDIA DriveWorks SDK, which implement Level 4 autonomous vehicles.
Affectiva, a software company that developed artificial intelligence to recognize user emotions, will explain how to measure driver emotions through computer vision and deep learning, enhance road safety, and implement more superior personalized transportation services.
Sessions related to HD maps, a key element in autonomous vehicle development, will also be conducted. Various map technologies that enhance autonomous driving capabilities, from "self-healing" maps for autonomous driving to 3D semantic maps that provide perception capabilities to autonomous vehicles, are expected to be introduced, with session presenters including Baidu, Civil Maps, HERE, TomTom, and Zenrin.
Additionally, sessions covering topics such as training simulations and artificial intelligence vehicle system certification will be held to accelerate autonomous vehicle development and provide more extensive and diverse data to autonomous vehicles that learn while driving.
관련뉴스
NVIDIA GRID Applied to Desktop Virtualization Infrastructure for Honda Vehicle Design
Nvidia announced that Nvidia GRID and Tesla M60 will be applied to Honda's graphics-accelerated desktop virtualization infrastructure. This project involves transitioning Honda's R&D division from physical workstations to a new vehicle design platform with improved mobility and efficiency, utilizing graphics-accelerated desktop virtualization infrastructure based on Nvidia GRID and Tesla M60.
2017.04.13by 김자영 기자
NVIDIA Hosts Deep Learning Day 2017 to Share Deep Learning and AI Technologies
NVIDIA will hold "NVIDIA DEEP LEARNING DAY 2017" beginning in Seoul on May 25th, followed by Daejeon on the 30th and Busan on June 1st, where it will share the latest deep learning and artificial intelligence trends and technologies. As interest in deep learning and artificial intelligence continues to grow and application fields expand progressively, NVIDIA is hosting "NVIDIA Deep Learning Day" for the second consecutive year to broaden the adoption of deep learning. Through NVIDIA Deep Learning Day, NVIDIA experts leading GPU-based deep learning technology will share the latest technologies and information on deep learning and artificial intelligence at the center of the Fourth Industrial Revolution with industry professionals, professors, and students, while providing practical training labs.
2017.04.17by 김지혜 기자
엔비디아, 페이스북과 협력해 AI 발전시켜 나갈것
엔비디아, 페이스북과 협력해 AI 딥 러닝 프레임워크 Caffe2 역량 강화 엔비디아는 페이스북이 오픈소스로 공개한 인공지능(AI) 딥 러닝 프레임워크인 카페2(Caffe2)와 함께 인공지능 발전을 위한 양사 협력의 결과를 발표했다. 페이스북은 넘쳐나는 데이터 관리를 돕는 인공지능 시스템을 개발하고 있다. AI 시스템으로 효율적인 커뮤니케이션을 돕는다. 개발자 및 연구원은 Caffe2를 활용해 대규모 분산 트레이닝 시나리오를 구성하고 사용자 디바이스용 머신러닝 애플리케이션을 개발할 수 있다. 모바일 상에서 인공지능 기반 서비스를 제공하기 위해서는 복잡한 데이터 처리 작업이 순식간에 진행되어야 한다. 이처럼 빠른 인공지능 서비스 처리는GPU 가속화 컴퓨팅과 가속화된 하드웨어의 전체 성능을 활용할 수 있는 딥 러닝 소프트웨어가 필요하다. ? 엔비디아와 페이스북은 Caffe2 딥 러닝 프레임워크 상에서의 공동 작업을 통해 인공지능 가속화를 제공하고 있다. 공동 엔지니어링을 통해 시작부터 끝까지 조정함으로써, 엔비디아 GPU 딥 러닝 플랫폼의 장점을 최대한 활용할 수 있도록 구성했다. Caffe2는 최신 엔비디아 딥 러닝 SDK 라이브러리인 cuDNN, cuBLAS, NCCL을 활용하여 고성능의 멀티 GPU 가속 트레이닝 및 추론을 제공한다. 이에 따라 사용자들은 Caffe2가 엔비디아 GPU 시스템에서 최적의 성능을 제공한다는 사실을 인지하고, 인공지능 기반 애플리케이션 개발에 집중할 수 있다. 본 협업의 일환으로 엔비디아 DGX-1 AI 슈퍼컴퓨터는 딥 러닝에 최적화된 소프트웨어 스택에 Caffe2를 제공하는 최초의 인공지능 시스템이 될 전망이다. DGX-1과 Caffe2 모두 높은 성능 및 빠른 트레이닝을 제공한다. 향후 고객들은 엔비디아 DGX-1 컨테이너 레지스트리(DGX-1 Container Registry)를 통해 DGX-1용 Caffe2를 이용할 수 있다. 그 동안 엔비디아는 엔비디아 딥 러닝 인스티튜트를 통해 전 세계 1만 여 명 이상의 개발자들이 프레임워크를 활용하여 다양한 지능형 애플리케이션 및 서비스용 뉴럴 네트워크 기반 머신러닝을 설계, 훈련, 구축할 수 있도록 지원해왔다. Caffe2 트레이닝은 현지시간 5월 8일부터 11일까지 미국 샌프란시스코 산호세에서 개최되는 GPU 테크놀로지 컨퍼런스에서 제공하는 엔비디아 딥 러닝 인스티튜트를 통해 만나볼 수 있다.
2017.04.21by 김자영 기자
A single misaligned light means 'defect'… Vehicle interior lighting now guarded by semiconductors and software
Automotive interior lighting is moving beyond its role as 'auxiliary lighting.' As it becomes a core element of user experience (UX) that expresses brand identity through color and animation while reflecting driver preferences, the LED interior ambient lighting market is growing rapidly. In connection with this, we met in advance with Choi Young-jae, Research Director at Dabo Corporation, who will present a technical webinar session on "Intelligent LED Interior Lighting Solutions Illuminating Future Vehicles," to hear about market trends and changes in design requirements.
2026.08.28by 명세환 기자
ST Microelectronics and NUS Establish Joint Research Center 'HELIX' for Edge AI
ST Microelectronics and the National University of Singapore have established the 'HELIX' joint research center aimed at implementing embedded AI in edge environments. Over a four-year period, the center will conduct research across the entire technology stack from algorithms to silicon, focusing on memory-centric architecture leveraging ST's 18nm FD-SOI technology and embedded PCM. The core objective is to implement intelligence directly in physical devices such as robots and drones while minimizing energy consumption.
2026.08.28by 배종인 기자
