相关论文: The End of Slow Networks: It's Time for a Redesign

In this paper, we conduct systematic measurement studies to show that the high memory bandwidth consumption of modern distributed applications can lead to a significant drop of network throughput and a large increase of tail latency in…

Developers of networked systems often work with low-level RDMA libraries to tailor network modules to take full advantage of offload capabilities offered by RDMA-capable network controllers. Because of the huge design space of networked…

Massive data centers are at the heart of the Internet. The rapid growth of Internet traffic and the abundance of rich data-driven applications have raised the need for enormous network bandwidth. Towards meeting this growing traffic demand,…

Over the years, communication speed of networks has increased from a few Kbps to several Mbps, as also the bandwidth demand, Communication Protocols, however have not improved to that extent. With the advent of Wavelength Division…

A massive number of devices are expected to fulfill the missions of sensing, processing and control in cyber-physical Internet-of-Things (IoT) systems with new applications and connectivity requirements. In this context, scarce spectrum…

Recently, IEEE 802.11ax Task Group has adapted OFDMA as a new technique for enabling multi-user transmission. It has been also decided that the scheduling duration should be same for all the users in a multi-user OFDMA so that the…

We present NetReduce, a novel RDMA-compatible in-network reduction architecture to accelerate distributed DNN training. Compared to existing designs, NetReduce maintains a reliable connection between end-hosts in the Ethernet and does not…

As modern AI workloads increasingly rely on heterogeneous accelerators, ensuring high-bandwidth and layout-flexible data movements between accelerator memories has become a pressing challenge. Direct Memory Access (DMA) engines promise high…

To fulfill the low latency requirements of today's applications, deployment of RDMA in datacenters has become prevalent over the recent years. However, the in-order delivery requirement of RDMAs prevents them from leveraging powerful…

Asynchronous Distributed Reinforcement Learning (DRL) can suffer from degraded convergence when model updates become stale, often the result of network congestion and packet loss during large-scale training. This work introduces a network…

RDMA is vital for efficient distributed training across datacenters, but millisecond-scale latencies complicate the design of its reliability layer. We show that depending on long-haul link characteristics, such as drop rate, distance and…

In this paper we present a new approach for distributed DBMSs called P4DB, that uses a programmable switch to accelerate OLTP workloads. The main idea of P4DB is that it implements a transaction processing engine on top of a P4-programmable…

Ultra-dense networks (UDNs) represent a transformative access architecture for upcoming sixth generation (6G) systems, poised to meet the surging demand for high data rates. Achieving precise synchronization across diverse base stations…

Distributed resource allocation (DRA) is fundamental to modern networked systems, spanning applications from economic dispatch in smart grids to CPU scheduling in data centers. Conventional DRA approaches require reliable communication, yet…

Support of real-time applications that impose strict requirements on packet loss ratio and latency is an essential feature of the next generation Wi-Fi networks. Initially introduced in the 802.11ax amendment to the Wi-Fi standard, uplink…

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…

Distributed deep learning (DDL) systems strongly depend on network performance. Current electronic packet switched (EPS) network architectures and technologies suffer from variable diameter topologies, low-bisection bandwidth and…

Limited memory bandwidth is a critical bottleneck in modern systems. 3D-stacked DRAM enables higher bandwidth by leveraging wider Through-Silicon-Via (TSV) channels, but today's systems cannot fully exploit them due to the limited internal…

Existing memory management techniques severely hinder efficient Large Language Model serving on accelerators constrained by poor random-access bandwidth.While static pre-allocation preserves memory contiguity,it incurs significant overhead…

Deep neural network (DNN) inference is increasingly being executed on mobile and embedded platforms due to low latency and better privacy. However, efficient deployment on these platforms is challenging due to the intensive computation and…