The GLU 3.0 solver, which is the GPU-based parallel sparse LU solver and was developed by Ph.D. student Shaoyi Peng, has been accepted to publish in IEEE Design and Test, which is  one of top journals in the electronic design automation area. This paper introduced the latest version of GLU solver, GLU 3.0, which features faster data dependency detection and dynamic GPU resource allocations for parallel sparse LU decomposition. GLU 3.0 represents the start of the art GPU-based sparse LU factorization techniques and has wide application in science and engineering field.  

The source codes, document and test examples of GLU 3.0 can be downloaded at https://intra.ece.ucr.edu/~stan/project/glu/glu_proj.htm

S. Peng and S. X.-D. Tan, "GLU3.0: Fast GPU-based parallel sparse LU factorization for circuit simulation", IEEE Design & Test, (in press). https://arxiv.org/abs/1908.00204.