No

Date	No	Name	ID	Topic	Presentation （50%）	Report （25%）	Roll Call （25%）	Final Score	Grade
1/22	1	孔昊然	225040481	GPU Communication Systems: Collective Communication Libraries	82
1/22	2	黄嘉铭	224040352	Research on Automated Code and Test Assertion Generation with LLMs	95
1/27	3	张馨元	225045037	End-to-End AI Inference Systems for Real-Time Healthcare	86
	4	彭一凡	225040521	Real-time System Optimization for ROS 2: Scheduling and Communication	96
	5	齐希贤	120090691	Beyond Algorithms: Hardware-Constrained Vector Search Databases	93
1/29	6	裴承轩	225040508	PD-Disaggregation in Large Language Models	92
1/29	7	陈张天艺	225040511	Tracing Operation System's Microkernel Journey and Its Performance Trade-offs	95
2/3	8	贾钊	225040505	Efficient Scheduling in Distributed OS	88
	9	张启航	119010434	When LLMs Become OS Operators: Rethinking Trust and Isolation	92
	10	陈俊颖	223040263	Evolution of Medical LLM Training Systems	94
2/5	11	庞威	225040490	Scheduling Deep Learning on GPU Clusters	98
2/5	12	陈启旭	120090643	Elastic Resource Provisioning in Cloud Platforms via Workload Prediction and Performance Modeling	90
3/5
3/5	13	毛宇	118010224	Data-Driven Control for Cloud Resource	95
3/10
	14	张文谦	225040483	From OS to Agentic OS	91
	15	李辉	224040351	Profile-Guided Optimization for Various Applications (OS kernel and data warehouse)	94
3/12
	16	周炫宁	225045030	Breaking the Memory Wall: FlashAttention and the Philosophy of IO-Aware Systems	93
	17	颜小川	225045041	From Hoare Logic to LLMs: Formal Verification and Generation of File Systems	95
3/17
	18	沈宇昊	225045038	From OS Paging to PagedAttention: Memory Management in Large Multimodal Models	92
	19	张书纶	225045020	LLM Routing: From Model Selection to Agentic Scheduling	95
3/19
	20	葛文韬	119010080	Memory-Efficient Large Model Training: From ZeRo to ZeRO-Infinity	95
	21	倪钦科	225045036	A Layered GPU Scheduling Architecture for Iterative Generative Workloads	90
3/24	22	廖欢	225040515	Three-Layer System Stack for Spoken Dialogue System	96
3/26
	23	房子皓	120090326	The Breakthrough Journey of Streaming Real-Time Speech Generation	92
	24	谭峙轩	225040506	The Evolution of AI Agents: From Memory Architecture to Self-Improving Agents	96
3/31	25	谢缘	224040374	KV Cache Management for Efficient LLM Serving	92
3/31	26	卢启晟	225040482	3D Human Motion Capture and Generation	93
4/2	27	王曼仪	225045034	From Benchmark to Agents: How SWE-bench Advanced LLM4Code	94
4/2	28	王楚娇	224045007	System Design for Large-Scale Reinforcement Learning Workloads	92
4/7	29	戴世成	225040523	Operating System Support for Large-Scale Graph Learning	97
4/7	30	陈骏安	225040494	The Evolution of Distributed LLM Training Systems	95
4/9	31	吴冠宗	224045015	System-Level Isolation for LLM Agent Safety	96
4/9	32	Juan Albert Wibowo	121040001	Securing Agent-Tool Interactions: A System-Level Approach	96
4/14	33	朱桐	225040538	OS-Inspired LLM Systems
4/14	34	张书源	225040535	From Docker to Kubernetes: A History of Container Management
4/16	35	郑博文	225040500	GPU Memory Optimization
4/16	36	李钺	225040518	From MicroVMs to Userspace Microkernels: Rethinking Isolation Boundaries in Cloud-Native Systems
4/21	37	刘效源	120040051	Autonomous Agent Systems for Discovery and Engineering
4/21	38	王匡	224040348	Branch Prediction Meets Token Prediction: OS Speculation Principles in Modern LLM Inference
4/23	39	李煜东	225040501	Mitigating System Latency in Streaming Voice Conversion: An OS-Level Perspective
4/23	40	胥瑶瑶	224040357	From GUI Agents to Computer-Use Agents: Foundation Models in Real Operating Systems
4/28	42	谢波涛	225045044	CXL-Enabled Memory Pooling: Redefining Memory Management in Distributed Systems
4/28	43	王瑞翔	225040514	Scheduling Asynchronous Inference in Robotic Systems