PoWER-BERT: Accelerating BERT inference via progressive word-vector eliminationSaurabh GoyalAnamitra R. Choudhuryet al.2020ICML 2020
Reducing Data Motion to Accelerate the Training of Deep Neural NetworksSicong ZhuangCristiano Malossiet al.2020arXiv
NeuNetS: An Automated Synthesis Engine for Neural Network DesignAtin SoodBenjamin Elderet al.2019arXiv
Incremental training of deep convolutional neural networksRoxana IstrateCristiano Malossiet al.2017AutoML 2017
Fast Exponential Computation on SIMD ArchitecturesCristiano MalossiYves Ineichenet al.2015HiPEAC 2015