Metraipolis › archive › dc_a544ef839da2

SymbolicLight V2: Hybrid Neuromorphic Architecture and Sparse Execution for Low-Energy Language Inference — arXiv:2609.09772

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.09772
License (as archived)
CC0-1.0
Archived
2026-10-16T07:53:20.088Z
SHA-256 of the archived text
b060b136689020a1f21308fbdbd0cb173bc0b13ce7d95d38474873707ec51f39
Stable URL of this page
https://metraipolis.com/documento/dc_a544ef839da2

How to cite

SymbolicLight V2: Hybrid Neuromorphic Architecture and Sparse Execution for Low-Energy Language Inference — arXiv:2609.09772 — arXiv (en), https://arxiv.org/abs/2609.09772 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a544ef839da2 (Metraipolis archive dc_a544ef839da2, sha256 b060b136689020a1f21308fbdbd0cb173bc0b13ce7d95d38474873707ec51f39)
BibTeX
@misc{dca544ef839da2,
  title = {SymbolicLight V2: Hybrid Neuromorphic Architecture and Sparse Execution for Low-Energy Language Inference — arXiv:2609.09772},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.09772},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a544ef839da2; sha256 b060b136689020a1f21308fbdbd0cb173bc0b13ce7d95d38474873707ec51f39},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.09772. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

SymbolicLight V2 combines sparse event computation with continuous-state processing in a hybrid neuromorphic language architecture. Extending V1's spike-gated dual paths, it adds graded signed events at further projections and softmax-free local attention. We implement the 194M-parameter model on an Alveo U50C FPGA using digital fixed-point arithmetic and on an ARM CPU using sparse integer execution. Across three same-checkpoint FPGA implementations at 175 MHz, active-row weight gathering and valid-state KV loading raise decode throughput from 474.6 to 643.2 tokens/s for a 32-token prefix and 128 outputs. Estimated gross card energy falls from 0.06087 to 0.04407 J per generated token, a 27.6% reduction. Complete-request energy, including prefill, falls by 24.4-27.7% across three prefix lengths. An independent idle split attributes 82.8% of gross card energy to loaded idle, explaining the benefit of shorter token latency. Against the recorded RTX 5090 compiled-FP32 baseline, integer FPGA execution uses 89.1% less estimated card energy during short-context decode; arithmetic precisions differ, and the GPU baseline is not the lowest-energy tested configuration. On four Cortex-A76 cores of a ROCK 5T, complete requests reach 65.4 tokens/s at 9.80 W and 0.151 J per generated token at the adapter's AC input. These results connect event sparsity to omitted computation and data movement. The mechanisms also support other dedicated V2 implementations: increasing throughput by a greater factor than active power lowers energy per generated token. Evaluation holds the deployed checkpoint fixed; its quality trails a same-budget dense control, so the results do not establish equal-quality efficiency.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-10 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.09772 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T07:24:00.896Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).