mlsanityData to 7 Oct 2026

A Diagonal Structured State Space Model on Loihi 2 for Efficient Streaming Sequence Processing

arXiv 2409.15022 ↗First posted 23 Sep 2024cs.LG cs.AI cs.ETcited by 6

On the map: State-space sequence models (S4, Mamba) → Core SSM architectures for language

Abstract (arXiv)

Deep State-Space Models (SSM) demonstrate state-of-the art performance on long-range sequence modeling tasks. While the recurrent structure of SSMs can be efficiently implemented as a convolution or as a parallel scan during training, recurrent token-by-token processing cannot currently be implemented efficiently on GPUs. Here, we demonstrate efficient token-by-token inference of the SSM S4D on Intel's Loihi 2 state-of-the-art neuromorphic processor. We compare this first ever neuromorphic-hardware implementation of an SSM on sMNIST, psMNIST, and sCIFAR to a recurrent and a convolutional implementation of S4D on Jetson Orin Nano (Jetson). While we find Jetson to perform better in an offline sample-by-sample based batched processing mode, Loihi 2 outperforms during token-by-token based processing, where it consumes 1000 times less energy with a 75 times lower latency and a 75 times higher throughput compared to the recurrent implementation of S4D on Jetson. This opens up new avenues towards efficient real-time streaming applications of SSMs.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Placed in Core SSM architectures for language, one of 10 directions in State-space sequence models (S4, Mamba). Builds on 34 of its 49 references that are in the map. First posted on arXiv 23 Sep 2024.

Builds on

34 of 49 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Mamba-360: Survey of State Space Models as Transformer Alternative for Long Sequence Modelling: Methods, Applications, and Challenges24 Apr
  2. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spaces1 Dec
  3. 2022Simplified State Space Layers for Sequence Modeling9 Aug
  4. 2022Diagonal State Spaces are as Effective as Structured State Spaces27 Mar
  5. 2021Efficiently Modeling Long Sequences with Structured State Spaces31 Oct

Other references · 29

  1. 2017Attention Is All You Needcited by 59,932
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2020End-to-End Object Detection with Transformerscited by 7,384
  4. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,611
  5. 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,031
  6. 2020Transformers are RNNs: Fast Autoregressive Transformers with Linear Attentioncited by 1,979
  7. 2019VideoBERT: A Joint Model for Video and Language Representation Learningcited by 702
  8. 2020Long Range Arena: A Benchmark for Efficient Transformerscited by 696
  9. 2017In-Datacenter Performance Analysis of a Tensor Processing Unitcited by 610
  10. 2021Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layerscited by 607
  11. 2020HiPPO: Recurrent Memory with Optimal Polynomial Projectionscited by 600
  12. 2022Efficiently Scaling Transformer Inferencecited by 529
17 more
  1. 2022On the Parameterization and Initialization of Diagonal State Space Modelscited by 418
  2. 2020The Hardware Lotterycited by 160
  3. 2022Liquid Structural State-Space Modelscited by 134
  4. 2024AI and Memory Wallcited by 89
  5. 2020Lipschitz Recurrent Neural Networkscited by 88
  6. 2023SpikeGPT: Generative Pre-trained Language Model with Spiking Neural Networkscited by 82
  7. 2018Learning Longer-term Dependencies in RNNs with Auxiliary Lossescited by 58
  8. 2021A Long Short-Term Memory for AI Applications in Spike-based Neuromorphic Hardwarecited by 32
  9. 2023Efficient Video and Audio processing with Loihi 2cited by 28
  10. 2023Green Edge AI: A Contemporary Surveycited by 24
  11. 2023Spiking Structured State Space Model for Monaural Speech Enhancementcited by 23
  12. 2023DYNAP-SE2: a scalable multi-core dynamic neuromorphic asynchronous spiking neural network processorcited by 20
  13. 2024SpikingSSMs: Learning Long Sequences with Sparse and Parallel Spiking State Space Modelscited by 19
  14. 2022Towards a General Purpose CNN for Long Range Dependencies in NDcited by 16
  15. 2024Q-S5: Towards Quantized State Space Modelscited by 9
  16. 2024Neuromorphic quadratic programming for efficient and scalable model predictive controlcited by 6
  17. 2024Solving QUBO on the Loihi 2 Neuromorphic Processorcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 5; 2026: 1.

6 papers.

  1. 2025State-Space Model Inspired Multiple-Input Multiple-Output Spiking Neuronscited by 4
  2. 2025Accelerating Linear Recurrent Neural Networks for the Edge with Unstructured Sparsitycited by 3
  3. 2025Quantizing Small-Scale State-Space Models for Edge AIcited by 3
  4. 2025Low-Bit Data Processing Using Multiple-Output Spiking Neurons with Non-linear Reset Feedbackcited by 1
  5. 2025QS4D: Quantization-aware training for efficient hardware deployment of structured state-space sequential modelscited by 0
  6. 2026Federated Learning of Spiking Neural Networks under Heterogeneous Temporal Resolutionscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Neuromorphic Principles for Efficient Large Language Models on Intel Loihi 2
  2. 2024IMSSA: Deploying modern state-space models on memristive in-memory compute hardware
  3. 2023Simple Hardware-Efficient Long Convolutions for Sequence Modeling
  4. 2026Density Field State Space Models: 1-Bit Distillation, Efficient Inference, and Knowledge Organization in Mamba-2
  5. 2025Compute-in-Memory Implementation of State Space Models for Event Sequence Processing
  6. 2026BaseRT: Advancing Best-in-Class LLM Inference with Apple M5 Neural Accelerators
  7. 2026Modern Neuromorphic AI: From Intra-Token to Inter-Token Processing
  8. 2025Advancing Intelligent Sequence Modeling: Evolution, Trade-offs, and Applications of State-Space Architectures from S4 to Mamba