Deep Neural Networks (DNNs) have emerged as the core enabler of many major
applications on mobile devices. To achieve high accuracy, DNN models have become …

Rewriting is a principled term transformation technique with uses across theorem proving
and compilation. In theorem proving, each rewrite is a proof step; in compilation, rewrites …

Convolution kernels are widely seen in deep learning workloads and are often responsible
for performance bottlenecks. Recent research has demonstrated that a direct convolution …

Improving the performance of deep neural networks (DNNs) is important to both the compiler
and neural architecture search (NAS) communities. Compilers apply program …

The convolution layer is the key building block in many neural network designs. Most high-
performance implementations of the convolution operation rely on GEMM (General Matrix …

Convolutions are the core operation of deep learning applications based on Convolutional
Neural Networks (CNNs). Current GPU architectures are highly efficient for training and …

Dilated convolutions are widely used to accomplish wide receptive fields while kee** the
resolution of feature maps in deep learning applications, such as semantic segmentation …

Graphics Processing Units (GPUs) are notoriously hard to optimize for manually. What is
needed are good automatic code generators and optimizers. Accelerate, Futhark and Lift …

Generating high-performance code for diverse hardware and application domains is
challenging. Functional array programming languages with patterns like map and reduce …

Equality saturation is a technique for implementing rewritedriven compiler optimizations by
efficiently representing many equivalent programs in so-called e-graphs. To improve …