Chris Forster

Chris Forster is a Senior CUDA Algorithms Software Engineer at NVIDIA, developing high-performance AI software applications. Prior to NVIDIA he was a Senior Member of Technical Staff at Sandia National Laboratories and Computational Physicist at SpaceX, developing large-scale multiphysics simulations and scientific software for the next generation of supercomputers and rocket engine design. He received his PhD from the Georgia Institute of Technology in Mechanical Engineering with an emphasis on physics simulation and HPC.

Posts by Chris Forster

AI / Deep Learning

Pretraining BERT with Layer-wise Adaptive Learning Rates

Training with larger batches is a straightforward way to scale training of deep neural networks to larger numbers of accelerators and reduce the training time. 10 MIN READ