NVIDIA 英伟达招聘深度学习性能架构师

初见你

NVIDIA 英伟达招聘深度学习性能架构师

我们是Glassdoor选出的“最佳工作场所”
我们是《人物》选出的“年度关爱公司”
我们是《财富》选出的“全球最受赞赏公司”

我们这里有无微不至的关爱
我们这里还有“评价最高的CEO”
......
我们，就是AI计算的引领者NVIDIA
......
梦想的机会就在眼前
你来不来？
如果你对职位感兴趣，请发简历至：Julietwang@careerintlinc.com
邮件标题：职位名+姓名+毕业时间+可到岗时间 +招聘信息来源

Deep Learning Performance Architect-上海/新竹
工作职责：
1.        针对架构和算法特征开发汇编级或者CUDA代码
2.        优化cuBlas、cuDNN、TensorRT的核心kernel；
3.        针对未来GPU架构开发原型代码，推进下一代架构的设计和优化；
基本要求：
1.        严谨的逻辑思维和分析能力
2.        较强编程能力、算法分析和实现
3.        有CUDA代码调优经验（或者SIMD等架构的调优经验）
4.        熟悉矩阵计算的优化和加速(优先)


Are you obsessed with performance?  Do you like to work at the intersection of
hardware and software?  Do you live and breathe deep learning?  NVIDIA is see
king world class programmers and performance architects who love to squeeze ou
t every cycle of performance from deep learning codes.  In this role, you will
write code that ships in our deep learning libraries, as well as guide the di
rection of our future GPU architectures.  This position offers the opportunity
to have real impact in a fast-moving, technology-focused company.
What you'll be doing:
•        Develop state of the art, performance critical code to accelerate deep
learning on NVIDIA's platforms.
•        Develop innovative HW, DSP, GPU and system architectures to extend the
state of the art in deep learning performance and efficiency
•        Analyze and prototype key deep learning and data analytics algorithms
and applications
•        Understand and analyze the interplay of hardware and software architec
tures on future algorithms and applications
•        Collaborate across the company to guide the direction of machine learn
ing, working with software, research and product teams
What we need to see:
•        MS or PhD in relevant discipline (CS, EE, Math)
•        Track record of optimizing code for performance on CPUs or GPUs, inclu
ding assembly or SIMD programming
•        Strong mathematical foundation in machine learning and deep learning
•        Experience working with deep learning frameworks like Caffe, TensorFlo
w and Torch
•        Strong programming skills in C, C++, Perl, or Python
•        Familiarity with GPU computing (CUDA, OpenCL, OpenACC) and HPC (MPI, O
penMP)
•        Strong background in computer architecture
•        Experience with matrix multiply and convolution algorithms