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 …

We present Polygeist, a new compilation flow that connects the MLIR compiler infrastructure
to cutting edge polyhedral optimization tools. It consists of a C and C++ frontend capable of …

Abstract Model Order Reduction (MOR) techniques play a crucial role in reducing the
computational complexity of high-dimensional mathematical models, enabling efficient …

Heterogeneous accelerators often disappoint. They provide the prospect of great
performance, but only deliver it when using vendor specific optimized libraries or domain …

Data-parallel computations, such as linear algebra routines and stencil computations,
constitute one of the most relevant classes in parallel computing, eg, due to their importance …

Reductions are common in scientific and data-crunching codes, and a typical source of
bottlenecks on massively parallel architectures such as GPUs. Reductions are memory …

The SAPFOR and DVM systems were primary designed to simplify the development of
parallel programs of scientific-technical calculations. SAPFOR is a software development …

Discovering and exploiting scalar reductions in programs has been studied for many years.
The discovery of more complex reduction operations has, however, received less attention …

MLIR is an emerging compiler infrastructure for modern hardware, but existing programs
cannot take advantage of MLIR's high-performance compilation if they are described in …

A Reduction–an accumulation over a set of values, using an associative and commutative
operator–is a common computation in many numerical computations, including scientific …