The demikernel datapath os architecture for microsecond-scale datacenter systems

I Zhang, A Raybuck, P Patel, K Olynyk… - Proceedings of the …, 2021 - dl.acm.org
Datacenter systems and I/O devices now run at single-digit microsecond latencies, requiring
ns-scale operating systems. Traditional kernel-based operating systems impose an …

{ODINFS}: Scaling {PM} performance with opportunistic delegation

D Zhou, Y Qian, V Gupta, Z Yang, C Min… - … USENIX Symposium on …, 2022 - usenix.org
Existing file systems for persistent memory (PM) exploit its byte-addressable non-volatile
access with low latency and high bandwidth. However, they do not utilize two unique PM …

Enabling high-performance and secure userspace nvm file systems with the trio architecture

D Zhou, V Aschenbrenner, T Lyu, J Zhang… - Proceedings of the 29th …, 2023 - dl.acm.org
Userspace library file systems (LibFSes) promise to unleash the performance potential of
non-volatile memory (NVM) by directly accessing it and enabling unprivileged applications …

Scale and performance in a filesystem semi-microkernel

J Liu, A Rebello, Y Dai, C Ye, S Kannan… - Proceedings of the …, 2021 - dl.acm.org
We present uFS, a user-level filesystem semi-microkernel. uFS takes advantage of a high-
performance storage development kit to realize a fully-functional, crash-consistent, highly …

Simurgh: a fully decentralized and secure NVMM user space file system

N Moti, F Schimmelpfennig, R Salkhordeh… - Proceedings of the …, 2021 - dl.acm.org
The availability of non-volatile main memory (NVMM) has started a new era for storage
systems and NVMM specific file systems can support extremely high data and metadata …

GPU-initiated on-demand high-throughput storage access in the BaM system architecture

Z Qureshi, VS Mailthody, I Gelado, S Min… - Proceedings of the 28th …, 2023 - dl.acm.org
Graphics Processing Units (GPUs) have traditionally relied on the host CPU to initiate
access to the data storage. This approach is well-suited for GPU applications with known …

{MT^ 2}: Memory Bandwidth Regulation on Hybrid {NVM/DRAM} Platforms

J Yi, B Dong, M Dong, R Tong, H Chen - 20th USENIX Conference on …, 2022 - usenix.org
Non-volatile memory (NVM) has emerged as a new memory media, resulting in a hybrid
NVM/DRAM configuration in typical servers. Memory-intensive applications competing for …

{FusionFS}: Fusing {I/O} Operations using {CISCOps} in Firmware File Systems

J Zhang, Y Ren, S Kannan - 20th USENIX Conference on File and …, 2022 - usenix.org
We present FusionFS, a direct-access firmware-level in-storage filesystem that exploits the
near-storage computational capability for fast I/O and data processing, consequently …

{ScalaCache}: Scalable {User-Space} Page Cache Management with {Software-Hardware} Coordination

L Peng, Y An, Y Zhou, C Wang, Q Li, C Cheng… - 2024 USENIX Annual …, 2024 - usenix.org
Due to the host-centric design principle, the existing page cache management suffers from
CPU consumption, communication costs, and garbage collection (GC) interference. To …

LabStor: A modular and extensible platform for developing high-performance, customized I/O stacks in userspace

L Logan, JC Garcia, J Lofstead, XH Sun… - … Conference for High …, 2022 - ieeexplore.ieee.org
Traditionally, I/O systems have been developed within the confines of a centralized OS
kernel. This led to monolithic and rigid storage systems that are limited by low development …