英伟达GPU Power Analysis Engineer
任职要求
• MS with 1-3 years of experience or PhD in related fields. • Basic understanding of concepts of energy consumption, estimation, and low power design. • Familiarity with Verilog and ASIC design principles, including knowledge of logic cells. • Good verbal/written English and interpersonal skills; much collaboration with design teams is expected. • Strong coding skills, preferably in Python, C++. • Ability to formulate and analyze algorithms, and comment on…
工作职责
• Use internally developed tools and industry standard pre-silicon gate-level and RTL power analysis tools, to help improve product power efficiency. • Develop and share best practices for performing pre-silicon power analysis, Enhance internal power tools and automate best practices • Perform comparative power analysis, to spot trends and anomalies, that warrant more scrutiny. • Interact with architects and RTL designers to help them interpret their power data and identify power bugs; drive them to implement fixes. • Select and run a wide variety of workloads for power analysis, Collaborate with performance and architecture teams to validate performance of the workloads • Prototype a new architectural feature in Verilog and analyze power.
• Use internally developed tools and industry standard pre-silicon gate-level and RTL power analysis tools, to help improve product power efficiency. • Develop and share best practices for performing pre-silicon power analysis, Enhance internal power tools and automate best practices • Perform comparative power analysis, to spot trends and anomalies, that warrant more scrutiny. • Interact with architects and RTL designers to help them interpret their power data and identify power bugs; drive them to implement fixes. • Select and run a wide variety of workloads for power analysis, Collaborate with performance and architecture teams to validate performance of the workloads • Prototype a new architectural feature in Verilog and analyze power.
• Use internally developed tools and industry standard pre-silicon gate-level and RTL power analysis tools, to help improve product power efficiency. • Develop and share best practices for performing pre-silicon power analysis, Enhance internal power tools and automate best practices • Perform comparative power analysis, to spot trends and anomalies, that warrant more scrutiny. • Interact with architects and RTL designers to help them interpret their power data and identify power bugs; drive them to implement fixes. • Select and run a wide variety of workloads for power analysis, Collaborate with performance and architecture teams to validate performance of the workloads • Prototype a new architectural feature in Verilog and analyze power.
• Use internally developed tools and industry standard pre-silicon gate-level and RTL power analysis tools, to help improve product power efficiency. • Develop and share best practices for performing pre-silicon power analysis, Enhance internal power tools and automate best practices • Perform comparative power analysis, to spot trends and anomalies, that warrant more scrutiny. • Interact with architects and RTL designers to help them interpret their power data and identify power bugs; drive them to implement fixes. • Select and run a wide variety of workloads for power analysis, Collaborate with performance and architecture teams to validate performance of the workloads • Prototype a new architectural feature in Verilog and analyze power.
工作职责 1、万卡级 GPU 调度系统建设： 参与大规模 GPU 集群调度系统建设，围绕 Quota、优先级、抢占、弹性伸缩、碎片整理、拓扑感知调度等能力提升资源效率。 2、训推统一调度： 面向大模型训练、后训练、推理服务等不同负载，设计训推统一调度、潮汐混部、在线离线协同和资源弹性策略。 3、资源利用率治理： 建设 GPU 资源利用率分析体系，基于真实负载数据识别低效资源、资源碎片、潮汐空闲和调度瓶颈。 4、LLMOps 平台融合： 参与构建面向大模型训练、微调、推理、部署全流程的 LLMOps 能力，与云原生平台深度融合，支撑大模型生产链路稳定高效落地。 5、集群稳定性建设： 与云原生、IDC、网络、存储和业务团队协作，提升大规模 AI 集群的故障恢复能力、资源周转效率和任务稳定性。 6、前沿技术探索： 持续关注 Kubernetes、Volcano、Kueue、Ray、GPU 虚拟化、弹性调度等相关技术，探索下一代 AI 资源调度系统。