René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Sparse * Dense Comparison with MKL","articleBody":"First of all, thank you so much for your interesting repo. Python-GraphBLAS has helped me a lot during my whole project. I just have a small question related to the comparison between Python-GraphBLAS's and MKL's matrix multiplication operation as follows:\r\n\r\nIn the Sparse * Dense scenario, I have run these 2 operations 10 times for benchmarking on the same input:\r\n```\r\nNumber of threads: 8\r\nX shape: (500091, 2381304)\r\nC shape: (100, 2381304)\r\n#nnz X: 1255206075\r\n#nnz C: 35204619\r\nDensity of X:  0.0010540255835157798\r\nDensity of C:  0.14783756714808358\r\n```\r\n\r\n```\r\nPython-GraphBLAS:\r\nX: csr\r\nC: fullc\r\n\r\nC = C.T\r\nXC \u003c\u003c 0\r\nXC \u003c\u003c XC(accum = gb.binary.plus, nthreads=nthreads) \u003c\u003c X.mxm(C)\r\n\r\nMean Runtime:  30.749253249168397\r\nStd Runtime:  0.1747113620436487\r\n```\r\nand\r\n```\r\nMKL:\r\nX: csr\r\nC: numpy.array(order=\"F\")\r\n\r\nC = C.T\r\nXC = sdm.dot_product_mkl(X, C)  \r\n\r\nMean Runtime:  17.70107755661011\r\nStd Runtime:  0.04377898424894914 \r\n```\r\nIt seems like the one conducted with MKL is more efficient. My question is whether I have used the optimal operation for Python-GraphBLAS? Is there any other way to conduct this operation in more efficient manner with Python-GraphBLAS?\r\n\r\nThank you so much in advance! I'm looking forward to hearing from you soon.","author":{"url":"https://github.com/khoinpd0411","@type":"Person","name":"khoinpd0411"},"datePublished":"2025-01-06T06:54:23.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":1},"url":"https://github.com/552/python-graphblas/issues/552"}