Summary

I am a developer and engineer with 10+ years of research and software development experience in the field of Computer Architecture, Virtualization and Distributed Systems in Academia and Industry. I have over 15 publications including top tier conferences and have contributed as external reviewer and technical program committees for more than 20 workshop and conferences covering multiple research areas - cloud, big data, system architecture and high performance computing. Experienced in software development of Distributed systems, Operating System and hypervisor based technologies ( e.g. VMware, Xen, KVM, Qemu, Linux Sandbox).  
I have a very strong understanding of  cloud technologies and new containers technologies - kubernetes, dockers and coreOS.
 
In past, I worked on next generation software defined storage solution for enterprise systems. at NVELO(Acquired by Samsung). Before  this, I worked as Senior Staff  at Huawei  Santa Clara.  My work domain include highly scalable virtual machine architectures, cloud computing, services and large scale data processing and analytics.  In addition, I worked on next generation data center which is based on large number of low power cores and contemporary interconnect  technologies. I was a lead architect for software layer of  light weight virtualization layer in typical data infrastructure for emerging data center architectures. In addition, I  explored cloud services, workloads (e.g. Hadoop) and highly scalable VM architectures. Mentored co-workers and initiated proposals.

In past, I worked as  Engineer in SoC Platform architecture team at Intel labs, Oregon. My work included QoS, Accelerator interfacing and heterogeneous computing in contemporary x86 based server, desktop and SoC architectures.  A typical O/S scheduler provides services differentiation to application ( that shares CPU between them) based on CPU execution time. I proposed control theory ( proportional integral controller ) based service differentiation based on  shared  resources such as caches, memory and power. The prototype was validated in typical multi-core architecture and representative workloads. In addition, I  explored the computation offload problem in typical x86 based SoC architecture - overhead in offloading task from CPU to IP block. I characterized typical overhead such as page table walk, system call , address space copy and proposed shared virtual memory between the hardware resources. On the other front, I worked on prototypes to characterize typical portability layer ( e.g. OpenCL ) for heterogeneous system. In addition, I worked on problem of intelligent scheduling of applications by O/S scheduler in completely heterogeneous systems consisting mix of low power and normal cores. For cluster environment, I have characterized emerging workloads (e.g. hadoop framework) for offloading heterogeneous jobs on backend clusters.

Before that, I did my PhD in Computer Information Science and Engineering from University of Florida under the guidance of Dr. Renato Figueiredo and Dr. Jose Fortes. I have been part of leading research labs - Advanced Computing Information and Science, NFS center of autonomic and Cloud Computing - as Graduate Research Assistant in Department of Electrical Engineering at University of Florida.  My dissertation, "Provisioning wide area environments through I/O interposition: The redirect-on-write file system and characterization of I/O overheads in a virtualized platform",  addressed I/O overheads in a  typical  para-virtualized stack ( e.g Xen Hypervisor) and I/O indirection over  wide area network. Broadly, I proposed proxy based shadow file system for write intensive workloads over wide are network - Redirect-on-Write File System.  Typical use cases for such scenarios are  non persistent containers over NFS and efficient diskless booting over wide area network.  On the other side, I have developed simulation based framework for overhead analysis of I/O intensive workload in virtualized environment.
 
During my PhD work, I worked twice as software research intern at Hardware abstraction lab and Many Core platform Architecture lab, Intel Corporation.  I worked on interesting research problem - What is the impact of microarchitectural resources such as caches and TLB on I/O networks in a virtualized environment ?  To address this, I worked on characterization of virtualized stack in  full system simulated environment.   This allowed me to research on impact of system resources on  typical para virtualized virtual machine. System performance analysis was done by both full system simulation ( e.g Simics) and performance tool chain such as oprofile on future Intel architectures. This work was published in top virtualization conference as first full system simulation of comprehensive study of  virtual machines for system resources (e.g. cache and TLBs). My internship was converted into a university grant for outstanding  intern performance. 
 
I have done my M. S. from the Mississippi State University in software Engineering under guidance of Dr. David Dampier ( Professor and Director of  forensics center at  MSU). My research included theoretical approach to change merging of abstract data types in object oriented applications. I have done my B. E in Electronics and Telecommunication from one of the oldest college in India, College of Engineering, Pune.
 
I have reviewed two books on virtualization  for virtual strategy magazines during my PhD work. In addition, I have reviewed book chapters for the authors of Operating System and Computer architecture books. I have contributed to open source bug report and patch. My efforts are duly acknowledged.
 
I love running, yoga and tracking. I have ran two half Marathons and many 10k and 5k runs in Florida, Portland and California over last five years. 
 
My contact email is vineet DOT chadha AT gmail DOT com.