Chieng-Chung Ho
Research Subjects
Chieng-Chung Ho has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
Bao-Kai Wang
1 shared publication- Da-Wei Chang
Department of Computer Science and Information Engineering, National Cheng Kung University, Tainan, Taiwan
1 shared publication I-Ting Lee
1 shared publicationLiang-Chi Chen
1 shared publicationWen Sheng Lim
1 shared publicationYu-Ming Chang
1 shared publication
Latest Publications
- Open Access
PIM or CXL-PIM? Understanding Architectural Trade-offs Through Large-Scale Benchmarking
Processing-in-memory (PIM) reduces data movement by executing near memory, but our large-scale characterization on real PIM hardware shows that end-to-end performance is often limited by disjoint host and device address spaces that force explicit staging transfers. In contrast, CXL-PIM provides a unified address space and cache-coherent access at the cost of higher access latency. These opposing interface models create workload-dependent tradeoffs that are not captured by small-scale studies. This work presents a side-by-side, large-scale comparison of PIM and CXL-PIM using measurements from real PIM hardware and trace-driven CXL modeling. We identify when unified-address access amortizes link latency enough to overcome transfer bottlenecks, and when tightly coupled PIM remains preferable. Our results reveal phase- and dataset-size regimes in which the relative ranking between the two architectures reverses, offering practical guidance for future near-memory system design.
Open AccessPIM or CXL-PIM? Understanding Architectural Trade-offs Through Large-Scale Benchmarking
Processing-in-memory (PIM) reduces data movement by executing near memory, but our large-scale characterization on real PIM hardware shows that end-to-end performance is often limited by disjoint host and device address spaces that force explicit staging transfers. In contrast, CXL-PIM provides a unified address space and cache-coherent access at the cost of higher access latency. These opposing interface models create workload-dependent tradeoffs that are not captured by small-scale studies. This work presents a side-by-side, large-scale comparison of PIM and CXL-PIM using measurements from real PIM hardware and trace-driven CXL modeling. We identify when unified-address access amortizes link latency enough to overcome transfer bottlenecks, and when tightly coupled PIM remains preferable. Our results reveal phase- and dataset-size regimes in which the relative ranking between the two architectures reverses, offering practical guidance for future near-memory system design.