Primal and dual block coordinate descent methods are iterative methods for solving regularized and unregularized optimization problems. Distributed-memory parallel implementations of these methods have become popular in analyzing large machine learning datasets. However, existing implementations communicate at every iteration which, on modern data center and supercomputing architectures, often dominates the cost of floating-point computation. Recent results on communication-avoiding Krylov subspace methods suggest that large speedups are possible by re-organizing iterative algorithms to avoid communication. We show how applying similar algorithmic transformations can lead to primal and dual block coordinate descent methods that only communicate every s iterations–where s is a tuning parameter–instead of every iteration for the regularized least-squares problem. We show that the communication-avoiding variants reduce the number of synchronizations by a factor of s on distributed-memory parallel machines without altering the convergence rate and attains strong scaling speedups of up to 6.1x on a Cray XC30 supercomputer.
Publications
Tags
2D
Accelerators
Algorithms
Architectures
Arrays
Big Data
Bootstrapping
C++
Cache Partitioning
Cancer
Careers
Chisel
Communication
Computer Architecture
CTF
DIABLO
Efficiency
Energy
FPGA
GAP
Gaussian Elimination
Genomics
GPU
Hardware
HLS
Lower Bounds
LU
Matrix Multiplication
Memory
Multicore
Oblivious
Open Space
OS
Parallelism
Parallel Reduction
Performance
PHANTOM
Processors
Python
Research Centers
RISC-V
SEJITS
Tall-Skinny QR
Technical Report
Test generation