The International Conference for High Performance Computing, Networking, Storage, and Analysis
The International Conference for High Performance Computing, Networking, Storage, and Analysis
Workshop: MEMO’25: International Workshop on Memory System Management and Optimization
Authors: Eric Tang and Tianyun Zhang (Carnegie Mellon University), William Bradford and Farzana Ahmed Siddique (University of Virginia), James C. Hoe (Carnegie Mellon University), Kevin Skadron (University of Virginia), and Franz Franchetti (Carnegie Mellon University)
Abstract: Data movement is a key bottleneck in applications such as machine learning and scientific computing. Some software techniques address this by computing on subsets of data but this still requires reading the entire dataset to determine the subset. We propose a hardware-software co-design approach for iterative methods centered around two operations--filtering and updating. We introduce a domain-specific language that supports these computational patterns to enable PIM programming. Since filter and update are simple pointwise operations, PIM hardware requires only limited compute capability.
In this work, we investigate gradient descent for an ill-conditioned convex optimization function using this approach and map it to a PIM architecture using the PIMEval architectural simulator. Filter load and update store operations sparsify the data set by 83% while requiring as few as 1.5x more iterations to converge compared to traditional gradient descent approaches, with a net reduction in data movement of 3.9x.
Back to MEMO’25: International Workshop on Memory System Management and Optimization Archive Listing Back to Full Workshop Archive Listing