Publications

Workshop GVDB24

Our paper titled “A Design Proposal for a Unified B-epsilon-Tree: Embracing NVM in Memory Hierarchies” has been accepted at the 35th GI-Workshop on Foundations of Databases, May 22-24, 2024, Herdecke, Germany) Abstract Non-volatile memory (NVM) represents a new class in the traditional storage hierarchy. The technologies in this class share characteristics of both primary and secondary storage; they provide latency that approaches that of DRAM, albeit moderately higher1, yet significantly lower than that of secondary storage devices, are addressable from cache lines, and, most importantly, offer persistence.

Transactions on Storage: Special Issue titled - Past, Present, and Future of Storage

Our manuscript entitled “NVM in Data Storage: A Post-Optane Future” which we had submitted as an abstract to the Transactions on Storage, Special Issue titled “Past, Present, and Future of Storage” has been reviewed. The reviewer(s) have recommended inviting us to submit a full paper, which will show up as a request for major revision of our abstract.

Workshop GVDB23

Our paper titled “Assessing Non-volatile Memory in Modern Heterogeneous Storage Landscape using aWrite-optimized Storage Stack” has been accepted at the 34th workshop on basics of database systems (Grundlage von Datenbanken) Abstract Non-Volatile memory (NVM), or persistent memory, is a promising and emerging storage technology that has not only disrupted the typical long-established memory hierarchy but also invalidated the proclaimed programming paradigm used in traditional database management systems and file systems. It bridges the gap between primary and secondary storage and, hence, shares the characteristics of both categories.

CHEOPS Workshop at EuroSys 2023

Our paper titled “Intelligent Data Migration Policies in a Write-Optimized Copy-on-Write Tiered Storage Stack” has been accepted at CHEOPS ‘23: Proceedings of the 3rd Workshop on Challenges and Opportunities of Efficient and Performant Storage Systems Abstract Modern storage stacks increasingly aim to intellegently optimize data accesses to improve latency, throughput, and energy consumption. Additionally, new storage setups become more heterogeneous, due to the introduction of new storage media, like NVM, and ever-increasing requirements regarding storage capacity and performance.