Concurrent Unrolled Skiplist

Platz, Kenneth; Mittal, Neeraj; Venkatesan, S.

doi:10.1109/icdcs.2019.00157

Cited by 6 publications

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, when each node stores a single element, the skip list exhibits no spatial locality. While it has been hypothesized that storing multiple elements per node (or "chunking") would improve performance, we are only aware of two published results [7,8]. The first does not seem to work efficiently for key types that are more complex than 32-bit integers, nor for maps with nontrivial value types.…”

Section: Introductionmentioning

confidence: 95%

Exploiting Locality in Scalable Ordered Maps

Rodriguez

Hassan

Spear

2020

Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques

View full text Add to dashboard Cite

This paper presents the skip vector, a novel high-performance concurrent data structure based on the skip list. Traversal is sped up by flattening the layers of the skip list into vectors, avoiding much of the costly pointer chasing that skip lists incur. The skip vector utilizes optimistic traversal with sequence locks, and hazard pointers for fast, memory-safe, concurrent access. In microbenchmark evaluation, we show that the skip vector offers excellent performance across a range of key ranges, thread counts, and operation mixes. CCS CONCEPTS • Computing methodologies → Concurrent algorithms.

show abstract

Section: Introductionmentioning

confidence: 95%