PRISM is the first PIM-based approach for sparse MTTKRP, reporting up to 2.64x speedup over CPU baselines on UPMEM hardware with heterogeneous CPU collaboration.
Sparsep: Towards efficient sparse matrix vector multiplication on real processing-in-memory architectures,
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
TokenStack's heterogeneous HBM-PIM design with base-die control and topology-aware KV placement delivers 1.62x higher geometric-mean token throughput and 1.70x SLO-compliant serving capacity than AttAcc while cutting per-token energy by 30-47%.
PIM-CACHE reduces mandatory coarse-grained transfers in UPMEM-style PIM by dynamically staging only non-redundant data via content-aware copy that exploits workload similarity.
citing papers explorer
No citing papers match the current filters.