Distributed out-of-memory NMF on CPU/GPU architectures
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States). Theoretical Division
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States). Computer, Computational, and Statistical Science Division
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States). HPC Division
We propose an efficient distributed out-of-memory implementation of the non-negative matrix factorization (NMF) algorithm for heterogeneous high-performance-computing systems. The proposed implementation is based on prior work on NMFk, which can perform automatic model selection and extract latent variables and patterns from data. In this work, we extend NMFk by adding support for dense and sparse matrix operation on multi-node, multi-GPU systems. The resulting algorithm is optimized for out-of-memory problems where the memory required to factorize a given matrix is greater than the available GPU memory. Memory complexity is reduced by batching/tiling strategies, and sparse and dense matrix operations are significantly accelerated with GPU cores (or tensor cores when available). Input/output latency associated with batch copies between host and device is hidden using CUDA streams to overlap data transfers and compute asynchronously, and latency associated with collective communications (both intra-node and inter-node) is reduced using optimized NVIDIA Collective Communication Library (NCCL) based communicators. Benchmark results show significant improvement, from 32X to 76x speedup, with the new implementation using GPUs over the CPU-based NMFk. Good weak scaling was demonstrated on up to 4096 multi-GPU cluster nodes with approximately 25,000 GPUs when decomposing a dense 340 Terabyte-size matrix and an 11 Exabyte-size sparse matrix of density 10-6.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE National Nuclear Security Administration (NNSA), Office of Defense Nuclear Nonproliferation; USDOE Laboratory Directed Research and Development (LDRD) Program
- Grant/Contract Number:
- 89233218CNA000001; AC52-06NA25396; 20190020DR
- OSTI ID:
- 2246858
- Report Number(s):
- LA-UR-23-33139
- Journal Information:
- Journal of Supercomputing, Vol. 80; ISSN 0920-8542
- Publisher:
- SpringerCopyright Statement
- Country of Publication:
- United States
- Language:
- English
Similar Records
A communication-avoiding 3D algorithm for sparse LU factorization on heterogeneous systems
Distributed non-negative RESCAL with automatic model selection for exascale data