TTLG - An Efficient Tensor Transposition Library for GPUs

This paper presents a Tensor Transposition Library for GPUs (TTLG). A distinguishing feature of TTLG is that it also includes a performance prediction model, which can be used by higher level optimizers that use tensor transposition. For example, tensor contractions are often implemented by using th...

Full description

Saved in:
Bibliographic Details
Main Authors: Vedurada, Jyothi, Suresh, Arjun, Rajam, Aravind Sukumaran, Kim, Jinsung, Hong, Changwan, Panyala, Ajay, Krishnamoorthy, Sriram, Nandivada, V. Krishna, Srivastava, Rohit Kumar, Sadayappan, P.
Format: Conference Proceeding
Language:English
Subjects:
GPU
Graphics processing units
High performance
Indexes
Instruction sets
Libraries
Taxonomy
Tensile stress
Tensor Transpose
Two dimensional displays
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper presents a Tensor Transposition Library for GPUs (TTLG). A distinguishing feature of TTLG is that it also includes a performance prediction model, which can be used by higher level optimizers that use tensor transposition. For example, tensor contractions are often implemented by using the TTGT (Transpose-Transpose-GEMM-Transpose) approach - transpose input tensors to a suitable layout and then use high-performance matrix multiplication followed by transposition of the result. The performance model is also used internally by TTLG for choosing among alternative kernels and/or slicing/blocking parameters for the transposition. TTLG is compared with current state-of-the-art alternatives for GPUs. Comparable or better transposition times for the "repeated-use" scenario and considerably better "single-use" performance are observed.
ISSN:1530-2075
DOI:10.1109/IPDPS.2018.00067