{"id":7,"identifier":"LOR-00007","url":"https://lifeofresearch.com/articles/7","title":"Neuromorphic Computing for On-Device LLM Inference: Why the Three-Layer Integration Gap Matters More Than the Algorithm Layer","abstract":"Spiking neural networks (SNNs) on neuromorphic hardware are routinely proposed as the path to energy-efficient on-device\ninference for language models. The framing usually treats this as a single technical question. It is not. There are three distinct\ntechnical layers — SNN algorithms for language tasks, neuromorphic hardware platforms, and the integration of the two into\nactual deployed systems — and each layer has a different state of progress, a different set of obstacles, and a different\nresearch community. This paper makes three claims. First, conflating the layers obscures where the genuine gaps lie: SNN-LLM\nalgorithm research has accelerated substantially with SpikeGPT, SpikeLLM, Sorbet, and recent spike-driven LLM\nconstructions, and neuromorphic hardware has matured with Loihi 2, IBM NorthPole, SpiNNaker 2, and BrainChip Akida —\nbut actual deployed SNN-LLM systems running on neuromorphic chips at the edge remain almost nonexistent. The integration\nlayer is the bottleneck, not the algorithms. Second, the energy-efficiency case for neuromorphic-LLM deployment must be\nmade against the right baseline — INT4 or sub-4-bit quantized models running on modern mobile NPUs — not against\nunquantized FP16 GPU baselines. The honest comparison narrows the advantage substantially and changes which workloads\nare worth pursuing. Third, the most plausible near-term wins are not as drop-in replacements for mobile-NPU LLM inference\nbut in specific niches: always-on low-rate token processing, sensor-fused language tasks where input is already event-based,\nand ultra-low-power deployment regimes where mobile NPUs do not operate. We propose a research agenda focused on the\nintegration gap and on rigorous baseline comparisons.","authors":[{"name":"Pranay Mahendrakar","affiliation":"","orcid":"0009-0003-7224-029X"}],"type":"article","license":"CC-BY-4.0","keywords":[],"publication_date":"2026-04-28","version":1,"views":7,"downloads":4,"created_at":"2026-09-22 17:33:40","doi":"10.5281/zenodo.19854579","source":"Zenodo","source_url":"https://zenodo.org/records/19854579","files":[{"id":7,"name":"Neuromorphic_OnDevice_LLM_Inference_Mahendrakar.pdf","size":92368,"mime":"application/pdf","downloads":4,"download_url":"https://lifeofresearch.com/articles/7/files/7/Neuromorphic_OnDevice_LLM_Inference_Mahendrakar.pdf"}]}