Related papers: Photonic Rails in ML Datacenters
In training of modern large natural language processing (NLP) models, it has become a common practice to split models using 3D parallelism to multiple GPUs. Such technique, however, suffers from a high overhead of inter-node communication.…
Optical data center networks (DCNs) are emerging as a promising solution for cloud infrastructure in the post-Moore's Law era, particularly with the advent of 'fast-switched' optical architectures capable of circuit reconfiguration at…
Coflow provides a key application-layer abstraction for capturing communication patterns, enabling the efficient coordination of parallel data flows to reduce job completion times in distributed systems. Modern data center networks (DCNs)…
Improving traffic management in case of perturbation is one of the main challenges in today's railway research. The great majority of the existing literature proposes approaches to make centralized decisions to minimize delay propagation.…
Mapping input signals to a high-dimensional space is a critical concept in various neuromorphic computing paradigms, including models such as Reservoir Computing (RC) and Extreme Learning Machines (ELM). We propose using commercially…
Offline Reinforcement Learning (RL) is structured to derive policies from static trajectory data without requiring real-time environment interactions. Recent studies have shown the feasibility of framing offline RL as a sequence modeling…
The advancement of artificial intelligence demands flexible multimodal data processing with high throughput and energy efficiency. Photonic integrated circuits (PIC) has demonstrated promising potentials in terms of low latency and low…
A new architectural paradigm, named, optical-computing-enabled network, is proposed as a potential evolution of the currently used optical-bypass framework. The main idea is to leverage the optical computing capabilities performed on…
Pipeline parallelism (PP) has become a standard technique for scaling large language model (LLM) training across multiple devices. However, despite recent progress in reducing memory consumption through activation offloading, existing…
The introduction of cloud data centres has opened new possibilities for the storage and processing of data, augmenting the limited capabilities of peripheral devices. Large data centres tend to be located away from the end users which…
Due to their intrinsic capabilities on parallel signal processing, optical neural networks (ONNs) have attracted extensive interests recently as a potential alternative to electronic artificial neural networks (ANNs) with reduced power…
With its unique parallel processing capability, optical neural network has shown low-power consumption in image recognition and speech processing. At present, the manufacturing technology of programmable photonic chip is not mature, and the…
Efficient traffic control (TSC) is essential for urban mobility, but traditional systems struggle to handle the complexity of real-world traffic. Multi-agent Reinforcement Learning (MARL) offers adaptive solutions, but online MARL requires…
High penetration from volatile renewable energy resources in the grid and the varying nature of loads raise the need for frequent line switching to ensure the efficient operation of electrical distribution networks. Operators must ensure…
Virtualisation first and cloud computing later has led to a consolidation of workload in data centres that also comprises latency-sensitive application domains such as High Performance Computing and telecommunication. These types of…
Oblivious routing has a long history in both the theory and practice of networking. In this work we initiate the formal study of oblivious routing in the context of reconfigurable networks, a new architecture that has recently come to the…
Modern deep learning relies nearly exclusively on dedicated electronic hardware accelerators. Photonic approaches, with low consumption and high operation speed, are increasingly considered for inference but, to date, remain mostly limited…
Machine learning (ML) is increasingly used in network data planes for advanced traffic analysis, but existing solutions (such as FlowLens, N3IC, BoS) still struggle to simultaneously achieve low latency, high throughput, and high accuracy.…
The decarbonisation of heavy-duty railway networks requires maximising the capacity of existing electrical infrastructure. Integrating heavy freight alongside fast passenger services exposes the hard physical limits of conventional…
Recently, Internet service providers (ISPs) have gained increased flexibility in how they configure their in-ground optical fiber into an IP network. This greater control has been made possible by (i) the maturation of software defined…