The Track on Simulation, Benchmarking, and Evaluation of Neuromorphic Systems invites high‑quality submissions that advance the methodologies, tools, and frameworks used to analyze, validate, and compare neuromorphic hardware and spiking neural network (SNN) platforms. As neuromorphic computing rapidly expands across edge intelligence, robotics, and large‑scale cognitive systems, rigorous evaluation is essential to ensure reproducibility, fairness, and meaningful progress. This track brings together researchers from architecture, circuits, computational neuroscience, machine learning, and systems engineering to establish the next generation of simulation and benchmarking practices for neuromorphic technologies.
We welcome original contributions including, but not limited to:
1. Simulation Frameworks and Tools
- Full‑system simulators for neuromorphic processors and SNN accelerators
- Mixed‑signal, analog, and device‑level simulation environments
- Multi‑scale simulation (device → circuit → architecture → system)
- Real‑time and hardware‑in‑the‑loop simulation platforms
- Co‑simulation of neuromorphic hardware with event‑based sensors
2. Benchmarking Methodologies
- Standardized benchmarks for SNNs, event‑driven processing, and neuromorphic workloads
- Evaluation suites for robotics, sensorimotor control, and edge intelligence
- Benchmarking of learning rules (STDP, R‑STDP, Hebbian, supervised SNN training)
- Metrics for latency, energy, throughput, accuracy, robustness, and scalability
- Comparative studies across digital, analog, mixed‑signal, and memristive systems
3. Performance Evaluation and Analysis
- Power, thermal, and reliability characterization of neuromorphic platforms
- Variability‑aware evaluation for analog and emerging‑device systems
- Communication and interconnect performance analysis for spike‑based traffic
- End‑to‑end evaluation of neuromorphic pipelines (sensing → processing → action)
- Profiling tools and instrumentation for neuromorphic hardware
4. Algorithms, Mapping, and Co‑Design Evaluation
- Evaluation of mapping, partitioning, and scheduling strategies for large SNNs
- Co‑design methodologies linking algorithms to hardware constraints
- Quantization, compression, and sparsity analysis for neuromorphic workloads
- Software frameworks enabling reproducible evaluation
5. Applications and Case Studies
- Real‑world deployments in robotics, autonomous systems, and edge AI
- Event‑based vision, audition, and tactile processing benchmarks
- Biomedical and brain–machine interface evaluation
- Comparative studies of neuromorphic vs. conventional AI systems
