Vignesh Vivekraja

About Vignesh Vivekraja

I am passionate about Deep Learning (DL) applications and algorithms, DL Hardware Acceleration, h/w-s/w Codesign and Processors.I work at Facebook on DL acceleration and co-design for AR/VR use cases.Before this I worked at Amazon on DL applications and its acceleration using AWS built custom instances called Inferentia.I previously worked at Nvidia, where I explored the interplay of hardware architectures and DL networks; and analyzed tradeoffs to achieve efficient acceleration of DL workloads.Prior to that, I worked at Intel as a micro-architect and logic design engineer for a Machine Learning accelerator IP, for audio/speech applications and the Intel CPU core.Back in grad school, my primary research focus was on ASIC architecture and circuits operating at sub-threshold voltage for ultra-low energy operation and hardware based security applications.

Experience

1. Senior Engineering Manager - Machine Learning

   Meta

   Apr 2020 — Present

   Technical lead and manage a team to architect software, hardware solutions for efficient deployment of ML algorithms for computer vision, imaging and large language models on ML Accelerators in Reality Labs products.Held various Tech Lead / IC roles.

Education

• Virginia Tech

  MS, Computer Engineering

  2008 — 2010

• National Institute of Technology Surat

  B Tech, Electronics Engineering

  2004 — 2008

Skills

• Application-Specific Integrated Circuits (Asic)
• Circuit Design
• Hardware Virtualization
• Logic Design
• Embedded Systems
• System on a Chip (Soc)
• Processors
• C
• Vlsi
• Cmos
• Deep Learning
• Asic
• Machine Learning Training
• Caffe
• Soc
• Machine Learning
• Rtl Design
• Systemc
• Computer Architecture
• Static Timing Analysis
• Hardware Accelerator
• Low Power Systems
• Hardware Architecture
• Cnn
• Verilog
• Python
• Rnn
• Tensor Flow
• High Level Synthesis