NTU教授招博士、博后职位 深度学习加速器 3D芯片架构建模 射频电路建模

SingularityKChenUID: 3993277

教授：Mohamed M. Sabry Aly (Asst Prof) https://scholar.google.com.sg/citations?user=eCo7XWkAAAAJ
教授背景：EPFL硕士、博士，斯坦福博后
邮箱：msabry@ntu.edu.sg


读博一年半了，再帮老板捞一手。明年8月入学的 PhD 名额还有最后一个，博后是都缺着。更新下组内情况以及我的感觉。
组内情况：一博后，仨（加上我）博士在读，其中一个博四快毕业了。目前依旧只有我一个中国人。跟新加坡科技局（ASTAR）合作非常多。

我的感受：一晃眼一年半了，2020年底中了第一篇 DAC，最近中了第二篇 DATA，手上第三个大项目做了大半年了。前两个中了论文的项目都是 ASTAR 的合作伙伴提供算法，我负责硬件设计。第一篇论文的芯片马上去流片了（主要是缺人所以拖了很久），手上的这个项目是 3D stacking + Chiplet，保底 40nm 工艺。入组后才渐渐明白老板的 funding 有多富 （保底 40nm 流片两次，最终要走 full mask的流片交差）。平常开会主要是跟 ASTAR 那边交流项目，有些事务型的内容才需要跟老板汇报。平常老板放养，不会 push 我，也放心让我慢慢搞项目（比如这个做了大半年的项目，我自己推了三版架构出来）。总的来说很舒服了。

感兴趣可以加我微信了解情况，跟这个账号名字一样。

以下是去年进组不久的时候发的帖子原文。
-------------------

组内情况：两博后，两博士在读，两研究硕士。目前就我一个中国人，日常沟通用英语。
我的感觉：进组内有4个月了，老板脾气是真好，有时还请喝咖啡；英语发音非常标准，没有沟通障碍。而且每季度都能流片，对于体系结构设计芯片的人来说，简直太棒了。希望我能赶上下一次流片。


NTU老板招博后～两个方向
1.Modeling and simulation of emerging 3D computing architectures
2.Simulation of wavefront computing systems

1.Modeling and simulation of emerging 3D computing architectures:
  This is part of a flagship AME programmatic project that aims to develop a new class of hardware accelerators for artificial intelligence, leveraging all knobs across the entire development stack (algorithm, architecture, circuit, and devices. The research fellow will be responsible for:

• Modeling computing systems that is enabled by 3D integration and new devices, at the architecture level. Candidate shold be capable of using architecture simulators such as Zsim or Gem5
• Analyzing and performing design space exploration for new classes of applications, not just AI
• perform architecture optimizations
• Thermal analysis and designs, using state of the art thermal simulators
• Investigate new computation paradigms for AI with these new devices
  

Candidates should hold a PhD degree and demonstrate ability to perform research through a number of high-quality publications. Expected date to join is Nov 2020 (project is till May 2023).


2.Simulation of wavefront computing systems.
The RF will work on on a project to develop a simulation and exploration framework to investigate new means of computing fabric enabled by ultrasonic waves.Roles of the RF will be:

• Development of compact model of the ultrasonic wave fabric and integrate it to an architectural simulation framework.
• Development of methodologies to estimate power and performance of this new compute fabric.
• Investigation of the memory connectivity to this new compute fabric.
• Supervise graduate student in modeling and in prototype demonstration.
  

Candidates should hold a PhD degree and demonstrate ability to perform research through a number of high-quality publications. Expected date to join is Nov 2020 (project is till Jan 2023).