• Corpus ID: 238253309

An Attempt to Generate Code for Symmetric Tensor Computations

  title={An Attempt to Generate Code for Symmetric Tensor Computations},
  author={Jessica Shi and Stephen Chou and Fredrik Kjolstad and Saman P. Amarasinghe},
Symmetric matrices, a frequently studied topic in linear algebra, can be extended to higher dimensions through symmetric tensors, which arise in domains such as computational physics and chemistry [5, 11]. The fundamental mathematical appeal of the study of symmetry also renders these tensors useful in contexts ranging from a rather beautiful equivalence with homogenous polynomials [3] to more concrete applications, including decompositions [9] and finding eigenvalues [8]. Knowing about a… 

Figures from this paper

Looplets: A Language For Structured Coiteration

This work shows how to abstract over array structures so that the compiler can generate code to coiterate over any combination of them, which enables new array formats, new iteration strategies, and new operations over structured data.



Efficiently Computing Tensor Eigenvalues on a GPU

This work describes how a GPU can be used to accelerate the computations of an application involving repeatedly solving the tensor eigenproblem for many small tensors, and presents an efficient implementation of Kolda and Mayo's algorithm.

The tensor algebra compiler

The first compiler technique to automatically generate kernels for any compound tensor algebra operation on dense and sparse tensors is introduced, which is competitive with best-in-class hand-optimized kernels in popular libraries, while supporting far more tensor operations.

A preliminary analysis of Cyclops Tensor Framework

This work details the automatic topology-aware mapping framework deployed by CTF, which maps tensors of any dimension and structure onto torus networks of any Dimension, and employs virtualization to provide completely general mapping support while maintaining perfect load balance.

Symmetric Tensors and Symmetric Tensor Rank

The notion of the generic symmetric rank is discussed, which, due to the work of Alexander and Hirschowitz, is now known for any values of dimension and order.

Performance Optimizations and Bounds for Sparse Symmetric Matrix-Multiple Vector Multiply

A performance model, which extends the authors' prior bounds for SpMV in the non-symmetric case, bounds performance by considering only the cost of memory operations, and using lower bounds on cache misses.

Eigenvalues of a real supersymmetric tensor

  • L. Qi
  • Mathematics
    J. Symb. Comput.
  • 2005

Exploiting Symmetry in Tensors for High Performance: Multiplication with Symmetric Tensors

This paper proposes a blocked data structure (blocked compact symmetric storage) wherein the tensor by blocks is considered and store only the unique blocks of a symmetric tensor.

New implementation of highlevel correlated methods using a general block tensor library for high-performance electronic structure calculations

  • 2013

Format abstraction for sparse tensor algebra compilers

An interface that describes formats in terms of their capabilities and properties is developed, and a modular code generator design makes it simple to add support for new tensor formats, and the performance of the generated code is competitive with hand-optimized implementations.