Welcome to Rutgers HPDA group's GitHub repository, where we're pushing the boundaries of data analytics with our advanced tensor operation customizations. Our work, encapsulated in numerous research papers, has led to the development of groundbreaking methodologies that significantly outperform established libraries like cuBLAS, CUTLASS, DGL, Gunrock and SuperLU. At the heart of our innovations is optimizing special tensor operations for accelerated data processing, ensuring faster, more efficient computations. Whether you're a researcher, data scientist, or enthusiast in the field of high-performance computing, our repository offers a treasure trove of resources and tools that stand at the forefront of computational efficiency. Dive in to explore our contributions and join us in shaping the future of data analytics.
We are still updating this repository, stay tuned!
Anil Gaihre
Santosh Pandey
Shiyang Chen
Lang Zhu
Haoshen Yang
Pinhuan Wang
Jiankun Jiang
[SC '15] Enterprise: Breadth-First Graph Traversal on GPUs [PDF]
[FAST'17] Graphene: Fine-Grained IO Management for Graph Computing [PDF]
[HPDC'19] XBFS: eXploring Runtime Optimizations for Breadth-First Search on GPUs [PDF]
[DAC '19] Dr. BFS: Data Centric Breadth-First Search on FPGAs [PDF]
[USENIX ATC '19] SIMD-X: Programming and Processing of Graph Algorithms on GPUs [PDF]
[SC '20] C-SAW: a framework for graph sampling and random walk on GPUs [PDF]
[TPDS '21] GSOFA: Scalable Sparse Symbolic LU Factorization on GPUs [PDF]
[TPDS '21] Trust: Triangle Counting Reloaded on GPUs[PDF]
[SC '21] Dr. Top-k: Delegate-Centric Top-k on GPUs[PDF]
[SC '21] E.T.: Re-Thinking Self-Attention for Transformer Models on GPUs[PDF]
[SC '23] Tango: Re-Thinking Quantization for Graph Neural Network Training on GPUs [PDF]