Low-Rank GEMM: Efficient Matrix Multiplication via Low-Rank Approximation with FP8 Acceleration
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912726396174336 |
|---|---|
| author | Metere, Alfredo |
| author_facet | Metere, Alfredo |
| contents | Large matrix multiplication is a cornerstone of modern machine learning workloads, yet traditional approaches suffer from cubic computational complexity (e.g., $\mathcal{O}(n^3)$ for a matrix of size $n\times n$). We present Low-Rank GEMM, a novel approach that leverages low-rank matrix approximations to achieve sub-quadratic complexity while maintaining hardware-accelerated performance through FP8 precision and intelligent kernel selection. On a NVIDIA RTX 4090, our implementation achieves up to 378 TFLOPS on matrices up to $N=20480$, providing 75\% memory savings and $7.8\times$ speedup over PyTorch FP32 for large matrices. The system automatically adapts to hardware capabilities, selecting optimal decomposition methods (SVD, randomized SVD) and precision levels based on matrix characteristics and available accelerators. Comprehensive benchmarking on NVIDIA RTX 4090 demonstrates that Low-Rank GEMM becomes the fastest approach for matrices $N\geq10240$, surpassing traditional cuBLAS implementations through memory bandwidth optimization rather than computational shortcuts. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_18674 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Low-Rank GEMM: Efficient Matrix Multiplication via Low-Rank Approximation with FP8 Acceleration Metere, Alfredo Performance Artificial Intelligence Distributed, Parallel, and Cluster Computing Machine Learning Large matrix multiplication is a cornerstone of modern machine learning workloads, yet traditional approaches suffer from cubic computational complexity (e.g., $\mathcal{O}(n^3)$ for a matrix of size $n\times n$). We present Low-Rank GEMM, a novel approach that leverages low-rank matrix approximations to achieve sub-quadratic complexity while maintaining hardware-accelerated performance through FP8 precision and intelligent kernel selection. On a NVIDIA RTX 4090, our implementation achieves up to 378 TFLOPS on matrices up to $N=20480$, providing 75\% memory savings and $7.8\times$ speedup over PyTorch FP32 for large matrices. The system automatically adapts to hardware capabilities, selecting optimal decomposition methods (SVD, randomized SVD) and precision levels based on matrix characteristics and available accelerators. Comprehensive benchmarking on NVIDIA RTX 4090 demonstrates that Low-Rank GEMM becomes the fastest approach for matrices $N\geq10240$, surpassing traditional cuBLAS implementations through memory bandwidth optimization rather than computational shortcuts. |
| title | Low-Rank GEMM: Efficient Matrix Multiplication via Low-Rank Approximation with FP8 Acceleration |
| topic | Performance Artificial Intelligence Distributed, Parallel, and Cluster Computing Machine Learning |
| url | https://arxiv.org/abs/2511.18674 |