B
Bulent Abali
Researcher at IBM
Publications - 119
Citations - 3623
Bulent Abali is an academic researcher from IBM. The author has contributed to research in topics: Memory management & Cache. The author has an hindex of 30, co-authored 119 publications receiving 3483 citations.
Papers
More filters
Patent
Efficient and accurate lookups of data by a stream processor using a hash table
TL;DR: In this article, a data lookup in a buffer triggered for a search string is triggered by a query, and a processor searches for a selection of pairs from among multiple pairs of a hash table read from at least one address hash of the query string and matching at least a data hash of query string.
Patent
Lightweight primary cache replacement scheme using associated cache
TL;DR: In this article, the authors provide a method to recover data from cache misses by retrieving the data from memory, the first cache storing at least a subset of data stored in the second cache.
Patent
Data compression accelerator methods, apparatus and design structure with improved resource utilization
TL;DR: In this article, the authors provide a design structure for a data compression accelerator system with a plurality of hardware data compression accelerators and a hash table shared by the plurality of accelerators.
Proceedings ArticleDOI
Touché: Towards Ideal and Efficient Cache Compression By Mitigating Tag Area Overheads
TL;DR: Touche as mentioned in this paper is a framework for storing multiple compressed blocks from arbitrary addresses within a cacheline without tag area overheads, which achieves an average speedup of 12% when compared to an uncompressed baseline.
Journal ArticleDOI
IBM POWER9 processor and system features for computing in the cognitive era
L. B. Arimilli,Bartholomew Blaner,B. C. Drerup,Charles F. Marino,Derek Edward Williams,E. N. Lais,F. A. Campisano,Guy Lynn Guthrie,Michael Stephen Floyd,R. B. Leavens,S. M. Willenborg,Ronald Nick Kalla,Bulent Abali +12 more
TL;DR: New POWER9 features for both system types are described to enable enterprise applications requiring large symmetric multiprocessor servers with large memory footprints, as well as one to two socket industry form-factor servers.