Results 171 to 180 of about 6,589 (215)

Swift: Rethinking RDMA Control Plane for Elastic Computing

arXiv.org
Elastic computing enables dynamic scaling to meet workload demands, and Remote Direct Memory Access (RDMA) enhances this by providing high-throughput, low-latency network communication. However, integrating RDMA into elastic computing remains a challenge,
Junxue Zhang, Han Tian, Xin-Hui Huang, Wenxue Li, Kaiqiang Xu, Dian Shen, Yong Wang, Kai Chen   +7 more
semanticscholar   +1 more source

Palladium: A DPU-enabled Multi-Tenant Serverless Cloud over Zero-copy Multi-node RDMA Fabrics

Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication
Serverless computing offers resource efficiency but suffers from a heavyweight data plane. We present Palladium, a DPU-offloaded serverless data plane enabling distributed zero-copy communication.
Shixiong Qi, Songyu Zhang, K. K. Ramakrishnan, D. Z. Tootaghaj, Hardik Soni, Puneet Sharma   +5 more
semanticscholar   +1 more source

PART: Pinning Avoidance in RDMA Technologies

2020 14th IEEE/ACM International Symposium on Networks-on-Chip (NOCS), 2020
State-of-the-art Remote Direct Memory Access (RDMA) engines pin communication buffers, complicating the programming model, limiting the memory utilization, and mandating a separate memory translation subsystem spanning the network interface card and the OS.
Antonis Psistakis, Nikos Chrysos, Fabien Chaix, Marios Asiminakis, Michalis Giannioudis, Pantelis Xirouchakis, Vassilis Papaefstathiou, Manolis Katevenis   +7 more
openaire   +1 more source

BiCC: Bilateral Congestion Control in Cross-datacenter RDMA Networks

IEEE Conference on Computer Communications
With the development of network-intensive applications like machine learning and cloud storage, there are two growing trends: (i) RDMA has been widely deployed to enhance underlying high-speed networks; (ii) applications are deployed on geographically ...
Zirui Wan, Jiao Zhang, Mingxuan Yu, Junwei Liu, Jun Yao, Xinghua Zhao, Tao Huang   +6 more
semanticscholar   +1 more source

Performance Isolation Anomalies in RDMA

Proceedings of the Workshop on Kernel-Bypass Networks, 2017
To meet the increasing throughput and latency demands of modern applications, many operators are rapidly deploying RDMA in their datacenters. At the same time, developers are re-designing their software to take advantage of RDMA's benefits for individual applications.
Yiwen Zhang, Juncheng Gu, Youngmoon Lee, Mosharaf Chowdhury, Kang G. Shin   +4 more
openaire   +1 more source

RDMA Point-to-Point Communication for LLM Systems

arXiv.org
Emerging Large Language Model (LLM) system patterns, such as disaggregated inference, Mixture-of-Experts (MoE) routing, and asynchronous reinforcement fine-tuning, require flexible point-to-point communication beyond simple collectives.
N. Licker, Kevin Hu, Vladimir Zaytsev, Lequn Chen   +3 more
semanticscholar   +1 more source

LEFT: LightwEight and FasT packet Reordering for RDMA

Asia-Pacific Workshop on Networking
RDMA, as a cutting-edge networking technology, has gained extensive adoption in large-scale data centers due to its exceptional characteristics, such as low and stable latency, high throughput and low CPU utilization.
Peihao Huang, Xin Zhang, Zhigang Chen, Can Liu, Guo Chen   +4 more
semanticscholar   +1 more source

RDMA-based dataflow management system

2017 IEEE First Ukraine Conference on Electrical and Computer Engineering (UKRCON), 2017
A system to control the data flow between detector sensors input streams for the data connected with a single event is elaborated. Obtained results show that a High Performance Computing cluster data exchange can be efficient AND fault-tolerant.
openaire   +1 more source

RDMA Load Balancing via Data Partition

2019 28th International Conference on Computer Communication and Networks (ICCCN), 2019
With the development of data center networks, traditional TCP/IP cannot support the demand in data centers. Remote Direct Memory Access (RDMA) technology could improve the performance of DCN significantly because of high throughput and low latency. However, load balancing is a key issue in RDMA which has not been solved distribute.
Yi Wang, Ya-nan Jiang, Qiufang Ma, Chen Tian, Bo Bai, Gong Zhang   +5 more
openaire   +1 more source

Switch-Assistant Loss Recovery for RDMA Transport Control

IEEE/ACM Transactions on Networking
RoCEv2 (RDMA over Converged Ethernet version 2) is the canonical method for deploying RDMA in Ethernet-based datacenters. Traditionally, RoCEv2 runs over the lossless network which is in turn achieved by enabling Priority Flow Control (PFC) within the ...
Qingkai Meng, Yiran Zhang, Shan Zhang, Zhiyuan Wang, Tong Zhang, Hongbin Luo, Fengyuan Ren   +6 more
semanticscholar   +1 more source