Director, Deep Learning Software
Santa Clara, Ca, Us
I founded and led the original deep learning software organization at NVIDIA. Was responsible initially for technical strategy and execution of NVIDIA's early deep learning software platform: cuDNN, cuBLAS, Digits, Image Compute Engine, GPU Rest Engine (precursor to Triton Inference Server), and what is now TensorRT.Previously: Director, CUDA Library and Software Solutions (10/2014 - 3/2015)Sr. Manager, CUDA Libraries and Algorithms (11/2012-10/2014)Responsible for all CUDA Library teams and other related software products.Manager, Emerging Applications (5/2011-11/2012)I founded and led the Emerging Applications group. Emerging Applications sought to develop enabling software technologies allowing end-users to access the power of GPU computing in a variety of application areas. Resulted in the AmgX and NVBIO products. This group was folded into the CUDA Libraries and Algorithms team in November 2012.Senior Research Scientist, NVIDIA Research (2/2008 - 4/2011)