Towards Topology-Aware Very Large-Scale Photonic AI Accelerators

• URL: http://arxiv.org/abs/2604.26966v1
• Date: Thu, 16 Apr 2026 22:38:48 GMT
• Title: Towards Topology-Aware Very Large-Scale Photonic AI Accelerators
• Abstract summary: Photonic accelerators offer a compelling alternative due to their inherent parallelism and high-speed matrix operations.<n>We present a scalable photonic accelerator architecture based on a modular scale-out paradigm using 4 X 4 photonic tensor core units.<n>We establish the Symmetric Grid Rule, demonstrating that symmetric topologies improve utilization by up to 6X while reducing memory access by over 40% compared to linear configurations.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The rapid growth of deep neural networks (DNNs) has exposed fundamental limitations in electronic accelerators, where data movement dominates energy consumption, commonly referred to as the memory wall. Photonic accelerators offer a compelling alternative due to their inherent parallelism and high-speed matrix operations. However, existing research largely focuses on device-level innovations, leaving system-level scalability insufficiently explored. In this paper, we present a scalable photonic accelerator architecture based on a modular scale-out paradigm using 4 X 4 photonic tensor core units. We perform a systematic architectural analysis that incorporates the practical scaling limits of photonic hardware, including insertion loss, fanout penalties, and laser power limits, which restrict monolithic photonic scaling. Through evaluation on representative DNN workloads (GoogleNet, ResNet-18, MobileNet, and AlphaGo Zero) with up to 1024 processing elements, we identify a topology-dominated scaling bottleneck in the photonic domain, termed the Utilization Wall, where performance is governed by grid topology rather than hardware size. We further establish the Symmetric Grid Rule, demonstrating that symmetric topologies improve utilization by up to 6X while reducing memory access by over 40% compared to linear configurations, which reveal that topology-aware scaling is essential for achieving energy-efficient and high-performance photonic AI accelerators.

Related papers

• UHD Low-Light Image Enhancement via Real-Time Enhancement Methods with Clifford Information Fusion [51.51707075741303]
  We propose a novel real-time UHD low-light enhancement network based on geometric feature fusion.<n>Our method achieves millisecond-level inference for 4K/8K images on a single consumer-grade device.
  arXiv  Detail & Related papers  (2026-04-10T13:47:04Z)
• Evolutionary Mapping of Neural Networks to Spatial Accelerators [64.13809409887254]
  We introduce the first evolutionary, hardware-in-the-loop mapping framework for neuromorphic accelerators.<n>We evaluate our approach on Intel Loihi 2, a representative spatial accelerator featuring 152 cores in a 2D mesh.<n>Our method achieves up to 35% reduction in total latency compared to default cores on two sparse multi-layer perceptron networks.
  arXiv  Detail & Related papers  (2026-02-04T16:28:08Z)
• LuxIA: A Lightweight Unitary matriX-based Framework Built on an Iterative Algorithm for Photonic Neural Network Training [36.03523572070848]
  Current state of the art PNN simulation tools face significant scalability challenges when training large-scale PNNs.<n>We introduce the Slicing method, an efficient transfer matrix computation approach compatible with back-propagation.<n>The Slicing method substantially reduces memory usage and execution time, enabling scalable simulation and training of large PNNs.
  arXiv  Detail & Related papers  (2025-12-24T17:31:51Z)
• What Is Next for LLMs? Next-Generation AI Computing Hardware Using Photonic Chips [34.52960723566363]
  Large language models (LLMs) are rapidly pushing the limits of contemporary computing hardware.<n>This review surveys emerging photonic hardware optimized for next-generation generative AI computing.
  arXiv  Detail & Related papers  (2025-05-09T05:19:14Z)
• Scalable Back-Propagation-Free Training of Optical Physics-Informed Neural Networks [12.726911225088443]
  Physics-informed neural networks (PINNs) have shown promise in solving partial differential equations (PDEs)<n>Photonic computing offers a potential solution to achieve this goal because of its ultra-high operation speed.<n>This paper proposes a completely back-propagation-free (BP-free) and highly salable framework for training real-size PINNs on silicon photonic platforms.
  arXiv  Detail & Related papers  (2025-02-17T23:45:23Z)
• Hardware-Efficient Photonic Tensor Core: Accelerating Deep Neural Networks with Structured Compression [15.665630650382226]
  We introduce a block-circulant photonic tensor core for a structure-compressed optical neural network (StrC-ONN) architecture.<n>This work explores a new pathway toward practical and scalable ONNs, highlighting a promising route to address future computational efficiency challenges.
  arXiv  Detail & Related papers  (2025-02-01T17:03:45Z)
• Streamlined optical training of large-scale modern deep learning architectures with direct feedback alignment [48.90869997343841]
  We experimentally implement a versatile and scalable training algorithm, called direct feedback alignment, on a hybrid electronic-photonic platform.<n>An optical processing unit performs large-scale random matrix multiplications, at speeds up to 1500 TeraOPS under 30 Watts of power.<n>We study the scaling of the training time, and demonstrate a potential advantage of our hybrid opto-electronic approach for ultra-deep and wide neural networks.
  arXiv  Detail & Related papers  (2024-09-01T12:48:47Z)
• Random resistive memory-based deep extreme point learning machine for unified visual processing [67.51600474104171]
  We propose a novel hardware-software co-design, random resistive memory-based deep extreme point learning machine (DEPLM) Our co-design system achieves huge energy efficiency improvements and training cost reduction when compared to conventional systems.
  arXiv  Detail & Related papers  (2023-12-14T09:46:16Z)
• All-optical graph representation learning using integrated diffractive photonic computing units [51.15389025760809]
  Photonic neural networks perform brain-inspired computations using photons instead of electrons. We propose an all-optical graph representation learning architecture, termed diffractive graph neural network (DGNN) We demonstrate the use of DGNN extracted features for node and graph-level classification tasks with benchmark databases and achieve superior performance.
  arXiv  Detail & Related papers  (2022-04-23T02:29:48Z)
• FPGA-optimized Hardware acceleration for Spiking Neural Networks [69.49429223251178]
  This work presents the development of a hardware accelerator for an SNN, with off-line training, applied to an image recognition task. The design targets a Xilinx Artix-7 FPGA, using in total around the 40% of the available hardware resources. It reduces the classification time by three orders of magnitude, with a small 4.5% impact on the accuracy, if compared to its software, full precision counterpart.
  arXiv  Detail & Related papers  (2022-01-18T13:59:22Z)
• Silicon photonic subspace neural chip for hardware-efficient deep learning [11.374005508708995]
  optical neural network (ONN) is a promising candidate for next-generation neurocomputing. We devise a hardware-efficient photonic subspace neural network architecture. We experimentally demonstrate our PSNN on a butterfly-style programmable silicon photonic integrated circuit.
  arXiv  Detail & Related papers  (2021-11-11T06:34:05Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.