开发工程师技术，能源
Developer Technology Engineer, Energy
我们的工作专注于视觉和AI计算的计算模型。在过去二十年中，NVIDIA在视觉计算领域处于领先地位，这是计算机图形学的艺术与科学，得益于我们发明的GPU。GPU也被证明在解决计算机科学中最复杂的问题方面非常有效。如今，NVIDIA的GPU模拟人类智能，运行深度学习算法，并作为能够感知和理解世界的计算机、机器人和自动驾驶汽车的“大脑”。我们希望与世界上最聪明的人一起壮大公司和团队，现在加入我们的团队比以往任何时候都更加激动人心！
NVIDIA正在寻找热情的、世界级的计算机科学家和工程师（Compute Developer Technology - DevTech），以在NVIDIA平台上加速能源模拟和AI工作流。您将专注于CUDA性能优化，针对地震处理（例如成像/反演流程）、油藏模拟、电网模拟器以及相关的HPC/AI生产工作流。您将与客户和合作伙伴的工程团队以及NVIDIA的产品和工程小组紧密合作，以在多GPU和多节点系统上实现可衡量的速度提升和可扩展的性能。
您将负责的工作：
- 使用CUDA内核、内存移动、并发性和端到端吞吐量对GPU加速的应用程序进行分析和优化。
- 推动整个技术栈的性能改进：
- CUDA C++内核优化、启动配置、内存层次结构、流/事件
- GPU库（如适用）：cuBLAS、cuFFT、cuSPARSE、cuSOLVER、NCCL
- 使用MPI+NCCL、CPU/GPU重叠、通信模式实现多GPU和多节点扩展
- 构建可重复的基准测试、性能报告和调优建议（包括前后对比、方法论、扩展曲线）
- 开发和维护参考实现、示例和/或补丁，以帮助客户代码实现性能和可移植性
- 支持客户项目（从POC到生产），包括调试正确性/性能问题，并就部署的最佳实践提供建议（容器、调度器、集群）
- 与内部团队合作，提交可操作的问题，验证修复方案，并根据能源领域的实际客户需求影响产品路线图
- 构建内部库和可重用代码，为未来的NVIDIA产品奠定基础
我们希望看到：
- 计算机科学、电子工程、物理、应用数学等相关领域的学士/硕士学历（或同等经验）
- 在Linux环境下具备扎实的C/C++和Python编程能力
- 具备CUDA编程和GPU性能优化的实际经验
查看英文原文
Our work at NVIDIA is dedicated towards a computing model focused on visual and AI computing. For two decades, NVIDIA has pioneered visual computing, the art and science of computer graphics, with our invention of the GPU. The GPU has also shown to be spectacularly effective at solving some of the most complex problems in computer science. Today, NVIDIA’s GPU simulates human intelligence, running deep learning algorithms and acting as the brain of computers, robots and self-driving cars that can perceive and understand the world. We are looking to grow our company and teams with the smartest people in the world and there has never been a more exciting time to join our team!
NVIDIAislookingforapassionate, world-class computer scientists and engineers (ComputeDeveloperTechnology - DevTech)toaccelerateEnergysimulationandAIworkflowsonNVIDIAplatforms.YouwillfocusonCUDAperformanceoptimizationforworkloadssuchasseismicprocessing(e.g.,imaging/inversionpipelines),reservoirsimulation,powergridsimulators,andrelatedHPC/AIproductionworkflows.Youwillworkhands-onwithcustomerandpartnerengineeringteamsaswellasNVIDIAproductandengineeringgroupstodelivermeasurablespeedupsandscalableperformanceonmulti-GPUandmulti-nodesystems.
What you will be doing:
- Profile,analyze,andoptimizeGPU-acceleratedapplicationswithemphasisonCUDAkernels,memorymovement,concurrency,andend-to-endthroughput.
- Driveperformanceimprovementsacrossthestack:
- CUDAC++kerneloptimization,launchconfiguration,memoryhierarchy,streams/events
- GPUlibraries(asapplicable):cuBLAS,cuFFT,cuSPARSE,cuSOLVER,NCCL
- Multi-GPUandmulti-nodescalingusingMPI+NCCL,CPU/GPUoverlap,communicationpatterns
- Buildreproduciblebenchmarks,performancereports,andtuningrecommendations(before/after,methodology,scalingcurves).
- Developandmaintainreferenceimplementations,examples,and/orpatchestocustomercodetoenableperformanceandportability.
- Supportcustomerengagements(POCstoproduction),includingdebuggingcorrectness/performanceissuesandadvisingonbestpracticesfordeployment(containers,schedulers,clusters).
- Collaboratewithinternalteamstofileactionableissues,validatefixes,andinfluenceroadmapbasedonrealcustomerrequirementsinEnergy.
- Build internal libraries and resusable code that would lead to future NVIDIA products.
What we needto see:
- BS/MS(orequivalentexperience)inCS/CE/EE/Physics/AppliedMathorrelatedfield.
- StrongprogrammingskillsinC/C++andPythononLinux.
- Hands-onexperiencewithCUDAprogrammingandGPUperformanceoptimizationconcepts.
- ExperienceprofilinganddebuggingperformanceusingtoolssuchasNVIDIANsightSystems/NsightCompute(orequivalent).
- Understandingofparallelcomputingandperformancefundamentals(vectorization,threading,NUMA,memorybandwidth/latency).
- Abilitytocommunicatetechnicalfindingsclearlytobothengineersandnon-engineers.
- 5+yearsrelevant experienceinGPU/HPCoptimization;strongtrackrecordofdeliveredspeedupsandscalingimprovements.
Ways to stand out from the crowd:
- Leadsperformancereviewswithcustomerstakeholders;createsreusableplaybooks/referencedesigns.Experience/Skills(typical)
- HPCexperiencewithMPI,distributedsystems,andmulti-nodeperformancetuning.
- Energy/HPCdomainexposure:
- Seismicprocessingpipelines,RTM/FWI-stylepatterns,FFT/stencil/linearalgebraheavycodes
- Reservoirsimulation(sparse/iterativesolvers),preconditioning,domaindecomposition
- Powergridsimulation/transientstability/optimizationworkflows
- ExperiencewithCI/perfregressiontesting,containerizedworkflows(Docker/Apptainer),andschedulers(Slurm).
- FamiliaritywithAIworkflowsusedalongsidesimulation(dataprep,training/inferenceintegration,pipelineperformance).
NVIDIA is widely considered to be one of the technology world’s most desirable employers. We have some of the most forward-thinking and hardworking people in the world working for us. If you're creative and autonomous, we want to hear from you!
Originally posted on Himalayas