Benchmarks were measured on local edge hardware using real heavy models including ResNet-152 and Llama 3.1 8B quantized. Implementation details are intentionally omitted. These figures come from a modest edge device on purpose: the same reuse layer runs on x86_64 servers and multi-GPU data-center hosts, where each avoided model call costs far more, so the savings scale up with your hardware.
ResNet-152 heavy GPU calls avoided. 9.96x speedup, 100% agreement at 10x recurrence.
LLM calls avoided. 10.03x speedup, 100% agreement at 10x recurrence.
Prompt-token reduction with 100% factual accuracy and 100% semantic agreement.
Total board energy reduction on a 5x recurrence workload.
About one fifth the energy of a prefix-cache-only path.
Llama 8B calls avoided at a conservative threshold, with 100% reuse precision.
Invented facts composed correctly. Corrections changed future answers immediately.
LLM calls on an unchanged rerun. A mid-pipeline change recomputed only affected steps.
Reuse runtime beside a 6.2 GB model. 40 KB to 3 MB per 1,000 reused answers.
The same reuse core powers NeuraFrame Embodied, and it is built so that learning over time does not slow it down.
In the Embodied test a simulated agent learned a six-situation scenario to 100% memory recall by the second epoch, then a fresh agent booted warm from the shipped memory pack and acted correctly with zero re-teaching.
The learned memory moves as one portable file. A robot boots knowing what it was taught in simulation, with no retraining and no model on board.
Verified exact answers and corrections are a hash lookup, constant time no matter how large the memory grows. Semantic memory is bounded, so scan time stays flat as it learns over months and years.