字节跳动Intelligent Computing Architecture and Operating System Researcher | 智能计算体系结构与操作系统研究员-筋斗云人才计划
任职要求
1. Academic Background： Got doctor degree, preferably majoring in software engineering, computer science, mathematics, artificial intelligence, or related fields. Strong capabilities in computer architecture; excellent coding skills, solid foundation in data structures and fundamental algorithms; proficiency in C/C++, Go, or Python. 2. Technical Proficiency Familiar with Linux operating systems, kernel, and network-related knowledge; prior development experience in these areas is preferred. Experience in performance optimization is a plus. 3. Problem-Solving Abilities Outstanding problem analysis and solving skills, with the ability to independently explore innovative solutions. 4. Collaboration and Communication Strong communication and teamwork skills, capable of collaborating with cross-functional teams to explore new technologies and drive technological advancements. 5. Professional Mindset Robust psychological resilience and adaptability, with the courage to tackle challenges and the ability to remain calm, composed, and flexible in complex situations. 1、获得博士学位，软件工程、计算机、数学、人工智能相关专业优先； 2、具备优秀的计算机体系结构能力，拥有出色的代码能力、扎实的数据结构和基础算法功底，熟练掌握C/C++或Go、Python； 3、熟悉Linux操作系统、内核、网络等领域相关知识，有相关开发经验者优先，有性能调优方案经验者优先； 4、具备出色的问题分析和解决能力，有自主探索解决方案的能力； 5、拥有良好的沟通协作能力，能和团队一起探索新技术，推进技术进步； 6、具备出色的心理素质与应变能力，面对困难勇于迎接挑战；面对复杂局面沉着、冷静、灵活。
工作职责
Team Introduction: The ByteDance System Department is responsible for the R&D, design, procurement, delivery, and operational management of the company's infrastructure ranging from chips to servers, operating systems, networks, CDNs, and data centers. It provides efficient, stable, and scalable infrastructure to support global services such as Douyin, Toutiao, and Volcano Engine. The current areas of operation include, but are not limited to: the design and construction of data centers, chip R&D, server development, network engineering, Volcano Engine's edge-cloud services, high-performance intelligent hardware development, intelligent delivery and operation of IDC resources, intelligent monitoring and early warning of hardware infrastructure, operating systems and kernels, virtualization technologies, compilation toolchains, supply chain management, and many other infrastructure-related areas. 团队介绍： 字节跳动系统部，负责字节跳动从芯片到服务器、操作系统、网络、CDN 、数据中心等基础设施的研发、设计、采购、交付与运营管理，为包含抖音、头条、火山引擎等全球业务提供高效、稳定、具备可扩展性的基础设施。部门当前业务开展包括不限于：数据中心设计建设、芯片研发、服务器研发、网络工程研发、火山引擎边缘云业务、高性能智能硬件研发、IDC资源智能交付与运维、硬件基础设施智能监控与预警、操作系统与内核、虚拟化技术、编译工具链、供应链管理等众多基础设施相关方向。 课题介绍： 在当今数字化时代，随着云计算、人工智能和大数据技术的深度融合，现代数据中心正面临着指数级增长的算力需求与现有计算架构效能瓶颈之间的突出矛盾。传统以通用CPU为核心的体系架构在应对多样化负载时，暴露出诸多问题。例如，内存子系统带宽与时延约束导致的 “内存墙” 效应持续加剧，异构计算单元间的数据搬运开销占比超过实际运算时间，安全可信执行环境带来的性能损耗超过 30%，单机柜算力密度提升受限于功耗密度阈值。与此同时，新兴工作负载如AI训练、图计算、时序数据库等呈现出动态异构特征，对计算架构提出了差异化需求，传统固定架构难以实现最优能效比。 操作系统作为计算机体系结构下重要的软件基础设施与核心技术，在这样的背景下也面临着巨大的挑战。随着计算需求的增长和技术的进步，传统的同构计算环境已无法满足日益复杂的计算任务。现代计算场景中，硬件架构呈现高度异构化，包括 CPU、GPU、FPGA、TPU、NPU、DPU 等，同时边缘计算、云计算形成分布式网络。传统操作系统难以高效管理跨节点、跨架构的资源。加之人工智能训练等场景需要低延迟、高吞吐、安全可信，动态弹性的分布式系统支持，这就要求操作系统具备跨异构资源的统一抽象与调度能力。学术界和工业界对下一代计算机操作系统在分布式微内核架构，异构资源调度算法，跨层优化与编译器支持，安全可信技术，虚拟化和 Serverless，AI 驱动操作系统内核优化以及操作系统内置 AI 推理引擎等方面展开了积极的探索和研究。 课题挑战： 方向一：体系化结构方向 1）负载特征与架构优化：建立数据中心动态负载特征建模框架，深入研究面向数据中心Workload的体系结构设计与优化方法，使系统能够更好地适应多样化的负载需求； 2）CPU核心架构创新：研究高性能低功耗CPU核心架构，积极探索超标量流水线与数据流引擎的融合设计，提升CPU的性能和能效； 3）新型内存层次构建：构建支持存算一体化的新型内存层次结构，研究基于3D堆叠技术的近存计算架构，重点突破高带宽互连拓扑优化、混合内存控制器设计、内存访问模式预测算法，解决 “内存墙” 等问题； 4）安全可信架构构建：构建安全可信计算架构，包括侧信道攻击防御的微架构级实现、侧信道安全架构、自动侧 / 隐蔽通道泄漏检测，确保系统在复杂环境下的安全性和完整性； 5）数据中心架构创新：探索整机柜级系统总线扩展，构建内存语义互联的新型数据中心架构，研究基于新型总线协议 (CXL/UALink) 的全局内存共享机制，提升数据中心的整体性能和资源利用率； 6）可靠性增强技术研究：研究可靠性增强技术，包括开发基于机器学习的故障预测模型，设计自修复的微架构容错机制，研究硬件静默故障检测，以及系统及IP可靠性特性研究和数据分析，保障系统的稳定运行。 方向二：操作系统方向 1）操作系统关键技术突破：突破传统单机操作系统存在的硬件资源利用局限、功能扩展与升级运维复杂、数据管理与共享不足、安全性与可靠性欠佳等问题。在计算高度异构以及计算环境分布化的情况下，从硬件到软件建立完整的信任链，保证整个系统的安全性和完整性。同时，有效地管理和协调多个节点间的通信、数据同步及故障恢复，设计高效的调度算法来匹配任务需求与最适合的计算资源，以最大化性能和效率。操作系统需要能够理解不同类型的计算任务，并能根据实时的工作负载动态调整资源分配，实现跨异构资源的统一抽象与调度； 2）跨领域知识融合：本课题需要融合OS、内核、算法、存储、虚拟化、网络、系统工程等多方面的跨领域知识和经验，以实现数据中心智能计算体系结构与操作系统的协同创新。
• Providing Ethernet and routing expertise to customers during project delivery to design, architect and test Ethernet networking solutions. • Work on multi-functional teams to provide Ethernet network expertise to server infrastructure builds, accelerated computing workloads and GPU enabled AI applications. • Crafting and evaluating DevOps automation scripts for network operations, crafting network architectures, and developing switch fabric configurations. • Implementing tasks related to network configuration and validation for data centers. • Create Methods of Procedure and deployment documents. • Use software tools to validate and monitor network performance.
We are aiming to leverage AI and other leading technology and dedicated to provide safe and reliable risk control capabilities behind payments. The core technologies include rule engines, model engines, intelligent algorithm models, etc., We are the leading platform with capabilities of high concurrent real-time risk calculations and massive big data analysis and processing. And as the core risk management tech platform for global payment business, we adopt a multi-center deployment architecture around the world. Here you may have the opportunity to learn more about and participate in the design and development of the following aspects: 1. Ultimate computing optimization at the millisecond level. 2. Behavior analysis and risk mining under massive data. 3. Global multi-center system architecture planning and high-availability solution design. 4. Participated in the design of R&D of risk control systems and big data platforms. You will also have the opportunity to explore the architectural design and implementation of cutting-edge technologies such as privacy computing and large models in risk control systems.
Team Introduction： The Search Team is primarily responsible for the innovation of search algorithm and architecture research and development (R&D) for products such as Douyin, Toutiao, and Xigua Video, as well as businesses like E-commerce and Local Services. We leverage cutting-edge machine learning technologies for end-to-end modeling and continuously push for breakthroughs. We also focus on the construction and performance optimization of distributed and machine learning systems — ranging from memory and disk optimization to innovations in index compression and exploration of recall and ranking algorithms — providing students with ample opportunities to grow and develop themselves. The main areas of work include: 1. Exploring Cutting-Edge NLP Technologies: From basic tasks like word segmentation and Named Entity Recognition (NER) to advanced business functions like text and multimodal pre-training, query analysis, and fundamental relevance modeling, we apply deep learning models throughout the pipeline where every detail presents a challenge. 2. Cross-Modal Matching Technologies: Applying deep learning techniques that combine Computer Vision (CV) and Natural Language Processing (NLP) in search, we aim to achieve powerful semantic understanding and retrieval capabilities for multimodal video search. 3. Large-Scale Streaming Machine Learning Technologies: Utilising large-scale machine learning to address recommendation challenges in search, making the search more personalized and intuitive in understanding user needs. 4. Architecture for data at the scale of hundreds of billions: Conducting in-depth research and innovation in all aspects, from large-scale offline computing and performance and scheduling optimization of distributed systems to building high-availability, high-throughput, and low-latency online services. 5. Recommendation Technologies: Leveraging ultra-large-scale machine learning to build industry-leading search recommendation systems and continuously explore and innovate in search recommendation technologies. 团队介绍： 字节跳动搜索团队主要负责抖音、今日头条、西瓜视频等产品以及电商、生活服务等业务的搜索算法创新和架构研发工作。我们使用前沿的机器学习技术进行端到端建模并不断创新突破，同时专注于分布式系统、机器学习系统的构建和性能优化，从内存、Disk等优化到索引压缩、召回、排序等算法的探索，充分给同学们提供成长自我的机会。 主要工作方向包括： 1、探索前沿的NLP技术：从基础的分词、NER，文本、多模态预训练，到业务上的Query分析、基础相关性等，全链路应用深度学习模型，每个细节都充满挑战； 2、跨模态匹配技术：在搜索中应用CV+NLP深度学习技术，实现多模态视频搜索强大的语义理解和检索能力； 3、大规模流式机器学习技术：应用大规模机器学习，解决搜索中的推荐问题，让搜索更加个性化更加懂你； 4、千亿级数据规模的架构：从大规模离线计算，分布式系统的性能、调度优化，到构建高可用、高吞吐和低延迟的在线服务的方方面面都有深入研究和创新； 5、推荐技术：基于超大规模机器学习技术，构建业界领先的搜索推荐系统，对搜索推荐技术进行探索和创新。 课题背景/目标： 随着大模型技术的快速发展，智能搜索领域迎来了新的机遇和挑战。传统搜索技术在面对海量数据、多模态信息以及用户复杂需求时，逐渐暴露出模型容量不足、语义理解能力有限、资源利用率低等问题。基于大模型的智能搜索构建旨在通过引入大模型技术，提升搜索系统的智能化水平，优化用户体验，并解决超大规模检索、复杂语义理解、资源高效利用等核心问题。具体目标包括： 1、探索大模型与排序算法的结合，提升个性化排序的精度和用户体验； 2、研究生成式检索算法，解决百亿乃至千亿级别候选库的超大规模检索问题； 3、利用大语言模型（LLM）提升复杂多义query的搜索满意度； 4、构建高性能、低资源消耗的大规模批流一体检索和计算系统，提升资源利用率。 课题挑战/必要性： 1、个性化排序的挑战：传统排序算法难以充分利用多模态信息（如文本、图像、视频等），且模型复杂度有限，无法满足用户对精准化和个性化搜索的需求； 2、超大规模检索的挑战：传统判别式模型在千亿级别候选库的检索中，面临模型容量不足、索引效率低下等问题，亟需新一代检索算法； 3、复杂query理解的挑战：用户搜索需求日益复杂，传统搜索引擎难以准确理解长难句、多义query的语义，导致搜索结果满意度低； 4、资源利用率的挑战：搜索系统存储和计算分离的架构导致资源利用率低，如何在保证性能的同时优化资源使用成为关键问题； 5、基于大模型的智能搜索构建是解决上述挑战的必要途径。通过引入大模型技术，可以显著提升搜索系统的语义理解能力、检索效率和资源利用率，从而为用户提供更精准、更高效的搜索体验。 课题内容： 1、个性化排序大模型研究； 2、超大规模生成式检索算法研究； 3、基于LLM提升复杂多义query的搜索满意度； 4、高性能大规模批流一体检索和计算系统。 涉及的研究方向：排序大模型、生成式检索与跨模态融合、大语言模型（LLM）与复杂query理解、高性能计算与存储架构。
About the Role TCL Industries seeks a visionary Embodied AI Architect to lead software architecture for next-generation home assistant robots. You will design intelligent, interactive systems, integrating perception, planning, navigation, manipulation, HRI, and emotional modules. Collaborate with multidisciplinary teams to drive robotics innovation. Key Responsibilities •System Architecture: •Design and oversee end-to-end software architecture for embodied AI in home robots. •Define interfaces and protocols for system components. •Manage sensor and camera data for informed AI decision-making. •Guide middleware, simulation, and tool selection. •Integration & Optimization: •Ensure seamless integration across hardware, AI algorithms, and cloud services. •Profile, debug, and optimize system performance. •Establish best practices for development, testing, and deployment. •Collaboration: •Work with hardware, data science, and product teams to align technical and product goals. •Cooperate with researchers to design and develop advanced algorithms for perception, reasoning, planning, human-robot interaction (HRI) and emotional intelligence •Strategic Input: •Contribute to long-term vision and planning. •Represent the lab at conferences and industry events.