【快手】异构计算平台架构师-AI/多媒体方向 at 快手 / NLPJob

注意该招聘信息在 100 天前已发布，有可能已失效

申请人

【快手】异构计算平台架构师-AI/多媒体方向

于快手 in 北京

1.异构计算平台架构师—人工智能方向
职位描述：
1. 您将从事最前沿的人工智能异构计算平台研发，并将有机会领导和参与构建最强大的机器学习计算平台，开发惠及广大用户并对行业产生深远影响的最新人工智能技术。
2. 针对公司业务快速发展的需求，规划设计与业务相匹配的AI异构产品构架，设计性能以及能耗匹配最佳的加速引擎，推动软硬件协同发展。
3. 全面评估市场上最新AI异构计算芯片，分析芯片性能和特点。创建深度学习的性能评估模型和基准测试benchmark，负责各类AI异构加速芯片的引入评估及性能测试流程。
4. 随时把握人工智能及深度学习的应用发展趋势，参与下一代机器学习产品设计。通过建模规划各类AI异构计算场景的解决方案。
职位要求：
1. 希望您熟悉主流深度学习框架(例如TensorFlow/Caffe/Pytorch)，对深度学习的training和inferencing都有深入了解，对AI芯片的现状及发展有独到的理解。
2. 优秀的编程能力，精通C/C++语言，掌握Linux软件开发环境，熟练运用脚本语言（例如python）
3. 较好的英文读写能力，EE/CS相关专业硕士以上学历，有相关工作经验；
4. 拥有以下特长者优先 a. 有CPU，GPU等功能性能建模经验 b. 有FPGA/ASIC芯片设计开发经验 c. 有Kernel/Driver，软硬件接口开发经验 d. 有边缘计算经验 e. 机器学习领域博士，并有高质量文章发表

Heterogeneous platform architect – Artificial Intelligence
Job Description：
1. You will be engaged in the development of cutting-edge artificial intelligence heterogeneous computing platforms, and will have the opportunity to lead and participate in the construction of the most powerful machine learning computing platform, develop the latest artificial intelligence technology that will benefit the users and have a profound impact on the industry.
2. In response to the rapid development of the company’s business, plan and design the heterogeneous product for AI applications that matches both the short and long term roadmap. Design the accelerators with best performance and lowest energy consumption; promote the coordinated development of software and hardware.
3. Evaluate comprehensively the latest AI heterogeneous computing chips, analyze their performance and features. Create the deep learning performance evaluation model and benchmark. Define the flow and spec of AI heterogeneous accelerator evaluation.
4. Gain insight into the application trends of artificial intelligence and deep learning; participate in the design of the next generation machine learning products. Plan the solutions for various AI heterogeneous computing scenarios through annalistic modeling.
Requirement：
1. You should be familiar with popular deep learning frameworks (such as TensorFlow/Caffe), have in-depth understanding of deep learning training and inferencing, and in-depth understanding of the current status and development of AI chips;
2. Excellent programming skills, proficient in C/C++ language, proficient in Linux software development environment, proficient in scripting language (such as python)
3. Good English reading and writing ability, EE/CS related major, Master degree or above, relevant work experience;
4. The following specialties are preferred a. Experience in functional modeling of CPU, GPU, etc. b. Experience in FPGA/ASIC chip design and development c. Kernel/driver, hardware/firmware interface d. Experience with edge computing e. Doctoral in machine learning, with high quality articles published

2.异构计算平台架构师—多媒体方向
职位描述：
1. 您将从事最前沿的多媒体异构计算平台研发，并将有机会领导和参与构建最强大的视频处理计算平台，为广大用户呈现多姿多彩的视频节目内容。
2. 针对公司业务快速发展的需求，规划设计性能与能耗匹配最佳的视频加速引擎，推动软硬件协同开发。
3. 全面评估市场上最新视频异构架构解决方案，分析他们的性能和特点，负责各类视频异构加速方案的引入评估及性能测试流程。
4. 随时把握视频压缩、视频传输、视频处理的应用发展趋势，参与下一代视频加速引擎的设计。并构建各类异构计算场景的应用方案。
职业要求：
1. 希望您熟悉常见的视频编解码算法标准，如H264，H265，VP9，AV1
等，对视频压缩实现方案的现状及发展有独到的理解。
2. 熟悉常用的视频处理算法，如去噪，帧率转换，去块，超分辨率等
3. 优秀的编程能力，精通C/C++语言，掌握Linux软件开发环境，熟练运用脚本语言（例如perl/python）
4. 较好的英文读写能力，EE/CS相关专业硕士以上学历，有相关工作经验；
5. 拥有以下特长者优先 a. 有多媒体芯片设计经验优先 b. 有CPU，GPU，视频压缩，视频处理等功能性能建模经验 c. 有FPGA/ASIC芯片设计开发经验 d. 有常用的开源开发平台，如ffmpeg，x264，x265等经验 e. 多媒体领域博士，并有高质量文章发表

Heterogeneous platform architect – Multimedia
Job Description：
1. You will be engaged in the development of cutting-edge multimedia heterogeneous computing platforms, and will have the opportunity to lead and participate in the construction of the most powerful video processing computing platform, which presents a variety of video content for the majority of users.
2. In response to the rapid development of the company’s business, plan the best video acceleration engine that matches the design performance and energy consumption; and promote the coordinated development of software and hardware.
3. Comprehensively evaluate the latest video heterogeneous architecture solutions on the market, analyze their performance and features, Define the flow and spec of various video heterogeneous acceleration solutions.
4. Tracking the development trends in video compression, video transmission, and video processing, and participate in the design of next-generation video acceleration engines. And build various applications for heterogeneous computing scenarios.
Requirement：
1. You should be familiar with common video compression standards, such as H264, H265, VP9, AV1 etc., have a unique understanding of the current status and development of video compression implementation.
2. Familiar with video processing algorithms such as denoising, frame rate conversion, deblocking, super resolution, etc.
3. Excellent programming skills, proficient in C/C++ language, master Linux software development environment, proficient in scripting language (eg perl/python)
4. Good English reading and writing ability, EE/CS related major or above, relevant work experience;
5. The following specialties are preferred a. Experience in multimedia chip design is preferred b. Experience in functional modeling of CPU, GPU, video compression, video processing, etc. c. Experience in FPGA/ASIC chip design and development, familiar with VHDL, Verilog d. Experience in the hardware/firmware development, I/O interface design, kernel or driver development e. Experience in open source development platforms, such as ffmpeg, x264, x265, etc. f. PhD in the field of multimedia, with high quality articles published