We propose to apply optical circuit switching to enable dynamic AllReduce reconfiguration for accelerating distributed machine learning. With simulated annealing-based optimization, theproposed AllReduce reconfiguration approach achieves 31% less average training time than existing solutions.
Publication Date: 6/29/2025 Event: OECC/PSC 2025 Reference: WC2-1: 1-3, 2025 Authors: Paul S. Westbrook, OFS Labs; Benyuan Zhu, OFS Labs; Kenneth S. Feder, OFS Labs; Zhou Shi, OFS Labs; Tristan Kremp, OFS Labs; Yaowen Li, NEC Laboratories America, Inc.; Ting Wang, NEC Laboratories America, Inc.; David J. DiGiovanni, OFS Labs Abstract: We discuss recent advances […]
We demonstrate a wavelength tunable Distributed-Vibration-Sensing over PON scheme using low-cost ITLA and Enhanced-Scattering-Fibers. Vibrations at frequency grids of 193.40THz and194.60THz in a PON with 1×16 splitter and 21 km feeder-fiber were successfully detected.
The recent proliferation of photorealistic images created by generative models has sparked both excitement and concern, as these images are increasingly indistinguishable from real ones to the human eye. While offering new creative and commercial possibilities, the potential for misuse, such as in misinformation and fraud, highlights the need for effective detection methods. Current detection approaches often rely on access to model weights or require extensive collections of real image datasets, limiting their scalability and practical application in real-world scenarios. In this work, we introduce a novel black-box detection framework that requires only API access, sidestepping the need for model weights or large auxiliary datasets. Our approach leverages a corrupt-and-recover strategy: by masking part of an image and assessing the models ability to reconstruct it, we measure the likelihood that the image was generated by the model itself. For black-box models that do not support masked-image inputs, we incorporate a cost-efficient surrogate model trained to align with the target models distribution, enhancing detection capability. Our framework demonstrates strong performance, outperforming baseline methods by 4.31% in mean average precision across eight diffusion model variant datasets.
Scene understanding systems analyze visual contexts by detecting objects, their attributes, and the interactions among them to provide a holistic interpretation. Understanding a scene requires analyzing multiple salient regions within a single video frame. Recently, Vision-Language Models (VLMs) have emerged as powerful tools for scene understanding, leveraging learned world knowledge to enable deployment without specialized training or fine-tuning. However, deploying VLMs in real-time applications is challenging due to their high computational and memory requirements, which limit processing throughput. We propose SimCache, a novel software-based caching mechanism that optimizes VLM-based scene understanding systems by reducing redundant computations. SimCache stores the embedding representation of a salient region and its detected activity, enabling reuse of VLM computations for similar regions in future frames. Specifically, SimCache exploits two types of redundancy: (1) temporal locality, reusing computations for similar regions across adjacent frames, and (2) semantic locality, reusing computations for visually distinct regions that represent the same activity at different times. SimCache includes a multi-tier cache architecture with specialized cache search and refinement policies to exploit redundancy efficiently and accurately. Experiments on action recognition datasets demonstrate that SimCache improves system throughput by up to 9.4× and reduces VLM computations by up to 24.4× with minimal accuracy loss.
Diffusion models (DMs) have proven to be effective in modeling high-dimensional distributions, leading to their widespread adoption for representing complex priors in Bayesian inverse problems (BIPs). However, current DM-based posterior sampling methods proposed for solving common BIPs rely on heuristic approximations to the generative process. To exploit the generative capability of DMs and avoid the usage of such approximations, we propose an ensemble-based algorithm that performs posterior sampling without the use of heuristic approximations. Our algorithm is motivated by existing works that combine DM-based methods with the sequential Monte Carlo (SMC) method. By examining how the prior evolves through the diffusion process encoded by the pre-trained score function, we derive a modified partial differential equation (PDE) governing the evolution of the corresponding posterior distribution. This PDE includes a modified diffusion term and a reweighting term, which can be simulated via stochastic weighted particle methods. Theoretically, we prove that the error between the true posterior distribution canbe bounded in terms of the training error of the pre-trained score function and the ]number of particles in the ensemble. Empirically, we validate our algorithm on several inverse problems in imaging to show that our method gives more accurate reconstructions compared to existing DM-based methods.
We present techniques to extend the sensing range in unrepeatered submarine cable systems by utilizing Enhanced-Scattering Fibre (ESF), large-area ultra-low-loss (ULL) fibre, and a digital Distributed Acoustic Sensing (DAS) interrogator. A DAS sensing range of up to 200.6 km has been achieved using 156km SCUBA125 fibre, followed by three segments of ESF. Additionally, we demonstrate long-range sensing capabilities and high-capacity data transmission over a 270.6 km unrepeatered submarine system, where DAS and 400G DWDM data transmission coexist. The impact of Distributed Raman Amplification (DRA) on sensing performance, and crosstalk between DAS and 400G DWDM channels in coexistence of DAS and unrepeatered transmission system are studied. Finally, we briefly discuss the potential application scenarios for monitoring undersea cables using ESFs.
We present a simple and accurate semi-analytical model predicting the gain of a single-stage erbium-doped fiber amplifier (EDFA) embedded with an unknown gain flattening filter (GFF). Characteristic wavelength-dependent gain coefficients and their scaling laws are extracted with a limited set of simple flat input spectrum measurements at variable temperatures and pump powers. Based on a black box approach, the proposed model provides a precise gain profile estimation of GFF-embedded EDFA for non-flat input spectra in variable temperature and pump power conditions. The accuracy of the presented methodology is validated on an extensive experimental dataset and compared with state-of-the-art gain models based on semi-analytic and solutions.
Subsea cables include a supervisory system that monitors the health of the amplifier pumps and fiber loss on per span basis. In some of the cables, the monitoring is achieved optically and passively using high-loss loop back paths and wavelength selective reflectors. By sending monitoring pulses through the supervisory channel and comparing the phases and polarizations of the returning pulses reflected by consecutive reflectors, dynamic disturbances affecting individual spans can be monitored on a per span basis. Such per-span phase monitoring techniques require high phase coherence compared to DAS systems since the spans are 10s of kms long compared to typical DAS resolution of meters. A time-frequency spread technique was demonstrated to limit the coherence length requirement, however the limits of its effectiveness was not quantified. In this paper we present a detailed analysis of the trade-off between implementation complexity and the phase noise tolerance for given span length by lab experiments.
Transformers, known for their attention mechanisms, have proven highly effective in focusing on critical elements within complex data. This feature can effectively be used to address the time-varying channels in wireless communication systems. In this work, we introduce a channel-aware adaptive framework for semantic communication, where different regions of the image are encoded and compressed based on their semantic content. By employing vision transformers, we interpret the attention mask as a measure of the semantic contents of the patches and dynamically categorize the patches to be compressed at various rates as a function of the instantaneous channel bandwidth. Our method enhances communication efficiency by adapting the encoding resolution to the contents relevance, ensuring that even in highly constrained environments, critical information is preserved. We evaluate the proposed adaptive transmission framework using the TinyImageNet dataset, measuring both reconstruction quality and accuracy. The results demonstrate that our approach maintains high semantic fidelity while optimizing bandwidth, providing an effective solution for transmitting multiresolution data in limited bandwidth conditions.
Princeton Office
4 Independence Way, Suite 200
Princeton, NJ 08540
+1 609-520-1555
San Jose Office
2033 Gateway Place, Suite 200
San Jose, CA 95110
+1 408-863-6007
NEC Laboratories America, Inc. (NEC Labs) is the US-based center for NEC Corporation’s global network of corporate research laboratories. Our diverse research groups collaborate with industry, academia and governments to provide disruptive solutions to complex problems. A leader in the integration of IT and network technologies with more than 100 years of expertise, NEC provides a combination of products and solutions that cross-utilize the company’s experience and global resources to meet the complex and ever-changing needs of its customers.