Parallel dynamic storage allocation algorithms
Arun Iyengar
Abstract
Arun Iyengar
Abstract
A dynamic storage allocator manages space for objects whose lifetimes are not known by the system at the time for their creation. This paper examines parallel dynamic storage allocation algorithms. Three new algorithms, multiple free list fit I, modified quick fit, and multiple free list fit II, are presented which outperform all previous algorithms we tested. The performances of the three algorithms differ when the percentage of requests for large blocks is high. Multiple free list fit I is the fastest algorithm when large blocks of many different sizes are requested. Multiple free list fit II is the fastest algorithm when large blocks of only a few different sizes are requested.>
OpenAlex reports 24 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A dynamic storage allocator manages space for objects whose lifetimes are not known by the system at the time for their creation. This paper examines parallel dynamic storage allocation algorithms. Three new algorithms, multiple free list fit I, modified quick fit, and multiple free list fit II, are presented which outperform all previous algorithms we tested. The performances of the three algorithms differ when the percentage of requests for large blocks is high. Multiple free list fit I is the fastest algorithm when large blocks of many different sizes are requested. Multiple free list fit II is the fastest algorithm when large blocks of only a few different sizes are requested.>
Key concepts: Allocator, Computer science, Algorithm, Linked list, Parallel computing, Theoretical computer science