【全职】NVIDIA招聘GPU Architects

nvidiahr

NVIDIA公司介绍

过去半个多世纪以来，人工智能一直是计算机科学家的梦想，而现在它将不再仅仅是科幻情景。在接下来的几年里，人工智能将给每一个行业带来变化。

在不久的将来，自动驾驶汽车将会减少交通拥堵、提升道路安全性。人工智能旅行社将会通过对你偏好的了解，从而为你的家庭度假安排好每一个细节。医学仪器也会读取并理解患者的 DNA 以检测癌症的早期迹象从而提前进行治疗。

曾几何时，发动机使人类变得更加强大并促成了第一次工业革命；而人工智能将会使人类变得更加智能并掀起下一场工业革命。

究竟是什么使得这一智能工业革命成为可能？

答案是基于GPU的深度学习这一全新的计算模式，它让计算机能够从数据中进行学习然后编写出人类无法写出的复杂软件。

事实证明，我们发明的 GPU 能够解决计算机科学领域中的一些最复杂的难题，其有效性令人难以置信。它起初被用作模拟人类想象力的引擎，用来为视频游戏和好莱坞电影大片呈现惊人的虚拟世界。现在，NVIDIA 的 GPU 能够模拟人类智能，可运行深度学习算法，并成为了那些可感知和了解世界的计算机、机器人以及自动驾驶汽车的大脑。

想象力与智能的无限扩展是我们奋斗终生的事业。

NVIDIA：人工智能计算公司

如果您对以下职位有兴趣，请将个人简历或者联系方式寄至echox@nvidia.com信箱。

Ø 1.SeniorGPU Graphics Architect 工作地点：上海

What you’ll be doing:

• Investigate and study state-of-the-art     real-time rendering techniques and the efficient support from GPU.
  
• Investigate and propose GPU graphics     architecture ideas to improve performance based on quantitative study of     existing and projected architectures.
  
• Develop performance simulation models and     simulation infrastructure.
  
• Develop performance testplan and tests for new     graphics units and architectural features.
  
• Test and debug on simulators, RTL and real     silicon.
  

What we need to see:

• Solid     knowledge of computer science, with the focus on computer graphics and/or     computer architecture.
  
• Good     understanding of GPU concept and pipeline, in terms of graphics processing     and parallel computing.
  " V# A) `! p  e! j% Y8 F
• Good     mastery of C++ language and at least 5 years of C/C++ development     experience.
  
• BSEE,     BSCSE, or equivalent required. MS or PhD is a plus.
  & u/ W$ M$ u& F1 |& T

Ways to stand out from thecrowd:

• Good     understanding of state-of-the-art rendering techniques and their usage of     GPU.
  0 A( u: l9 d2 H& k  D
• Experience     of performance modelling, profiling and analysis is a plus.
  
• Experience     with trace-driven and execution-driven simulation model development is a     plus.
  3 [* G% |  z: W- B0 }' m. H
• Candidates     with experience of GPU graphics unit design and/or RTL module development     are preferred.
  

Ø 2.Senior GPU Performance Architect 工作地点：上海

Responsibilities:

TheNVIDIA Architecture group is looking for world class programmers to develop thecore infrastructure for performance analysis, debugging, and tuning of largescale general purpose Graphics chips.  

Thisinfrastructure provides insights to our architects, driver developers, andDeveloper Relations teams for delivering the most polished and performantgaming experience.

Akey part of NVIDIA’s strength is our proprietary internal tools for thevisualization, analysis, and debug and verification of tests and applicationson various functional and performance simulations of future chips.

Weare continuously pushing the envelope and needing ever more sophisticated toolsto enable the next generation of chips.   These tools are used byhundreds of engineers worldwide.

Inthis role you will play a critical part in every stage of both the developmentand post-production tuning of a graphics chip for the most modern games.

Youwill work closely with other project members and users to deliver a suite ofanalysis tools that especially targets the unique properties of next generationgraphics APIs.  

Youwill become a customer of your own tools and learn to tune the performance ofthe most advanced GPUs in the world.  You will learn and greatly improvethe daily workflows of the world’s top chip modelers and designers.

Whatwe need to see:

·        Solidbackground in mathematics, algorithms and data structures required.

·        Experiencein developing web-based user interfaces and applications, by using Java &JavaScript or PHP

o   Knowledge of best practices for data visualizationa plus.

·        Experiencein real time 3D rendering or ray tracing, including DirectX or OpenGL API

o   ​Knowledge of next generation Graphics API’s(DX12/Vulkan) a plus

·        Astrong team player; self-motivated and good attitude is a must.

Ø 3.GPU SM Architect 工作地点：上海

Weare now looking for a GPU SM ARCHITECT – PERFORMANCE /FUNCTIONAL MODELING,VALIDATION AND ANALYSIS of SHADER UNIT

What you’ll be doing:

·        Investigateand propose architecture ideas based on quantitative study of existing andprojected SM architecture.

·        Developperformance and functional simulation models.

·        Developperformance and functional test plan and tests to validate new SM architecturaland features.

·        Testand debug on simulators, RTL and real silicon.

What we need to see:

·        Indepth knowledge of computer architecture, with good understanding of modern ISAand microprocessor implementation techniques.

·        Goodunderstanding of GPU concept and pipeline, in terms of Graphics processing andparallel compute.

·        Goodmastery of C++ language and at least 3 years of C/C++ development experience.

·        Experienceof performance/functional modelling, profiling and analysis is a plus.

·        Experiencewith trace-driven and execution-driven simulation model development is a plus.
BSEE, BSCSE, or equivalent required. MS or PhD is a plus.

·        Candidateswith GPU shader unit design, microprocessor design, and CPU/GPU performanceanalysis experience are preferred.

' I6 Q/ n& h0 p

Ways to stand out from thecrowd:

·        Candidateswith GPU shader unit design, microprocessor design, and CPU/GPU performanceanalysis experience are preferred.

Ø 4.Compute Architect 工作地点：上海、北京

What you’ll be doing:

·        Enhance GPU architectures to extend the stateof the art in performance, efficiency, programmability and reliability

·        Analyze, tune and prototype key applicationsfor new GPU architectures

·        Build model to predict performance, power andreliability on future architecture

·        Design, and develop tools to analyze,simulate, validate, and verify application performance, and energy consumption

·        Collaborate across the company to guide thedirection of GPU computing, working with software, research and product teams

What we need to see:

·        MS Degree in relevant discipline (CS, EE,Math). PhD helpful

·        Strong programming skills in C or C++

·        Strong background in computer architecture, parallelprocessing, signal processing and/or high performance computing

Ways to stand out from the crowd:

·        Big plus with deep learning, computer visionand HPC domain

·        Experience in characterizing and modelingsystem-level performance or power, executing comparison studies, anddocumenting and publishing results

·        Optional/Plus - Strong programming skills inCUDA, Perl, OpenMP, MPI or Python

Ø 5.System Design Software Engineer 工作地点：上海

We are now looking for a System DesignSoftware Engineer.

Whatyou'll be doing:

•           Weare looking for world class engineers to design, model, analyze and verify nextgenerations of GPU architecture.

•           Thecandidates will work with a group of architects to design and developproprietary internal tools for the visualization, analysis, and debug andverification of tests and applications on various functional and performancesimulations of future chips.

•           Thecandidates will have opportunities to get involved in cutting-edge GPU macro-and micro-architecture design, verification and optimization, including portingcommercial applications to test benches, identifying performance hotspots anddata mining for performance analysis.

Whatwe need to see:

•           Bachelor'sDegree or higher majoring in CS/EE/Mathematics or relevant fields.

•           Solidcomputer science background

•           StrongC/C++ programming ability.

•           ExcellentEnglish writing for engineering documentation, English oral well enough toattend meetings.

•           Experiencein the following areas is a plus:

•           Scriptinglanguage (Perl, Python, Ruby) experience is a plus.

•           3Dgraphics (D3D or OpenGL) application development.

•           Parallelcomputing/CUDA/OpenCL/HPC development.

•           Microprocessorarchitecture design & verification.

•           Systemlevel programming experience in OS, compiler, driver, tools, virtual memorysystem, etc.

•           Multimedia(video, image processing, visualization) application development

/ l! Q$ f6 d7 L( M3 l+ D