There is an increasing need to bring machine learning to a wide diversity of hardware
devices. Current frameworks rely on vendor-specific operator libraries and optimize for a …

We introduce a learning-based framework to optimize tensor programs for deep learning
workloads. Efficient implementations of tensor operators, such as matrix multiplication and …

This paper introduces Tiramisu, a polyhedral framework designed to generate high
performance code for multiple platforms including multicores, GPUs, and distributed …

Scalable frameworks, such as TensorFlow, MXNet, Caffe, and PyTorch drive the current
popularity and utility of deep learning. However, these frameworks are optimized for a …

DNN layers are multi-dimensional loops that can be ordered, tiled, and scheduled in myriad
ways across space and time on DNN accelerators. Each of these choices is called a …

Transformers are one of the most important machine learning workloads today. Training one
is a very compute-intensive task, often taking days or weeks, and significant attention has …

DNN accelerators provide efficiency by leveraging reuse of activations/weights/outputs
during the DNN computations to reduce data movement from DRAM to the chip. The reuse is …

Futhark is a purely functional data-parallel array language that offers a machine-neutral
programming model and an optimising compiler that generates OpenCL code for GPUs …

Graph-based intermediate representations (IRs) are widely used for powerful compiler
optimizations, either interprocedurally in pure functional languages, or intraprocedurally in …

This work presents topology optimisation implementations for linear elastic compliance
minimisation in three dimensions, accelerated using Graphics Processing Units (GPUs) …