arXiv:2603.13606v3 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) architectures have become essential for scaling large language models, driving the development of specialized device-initiated communication libraries such as DeepEP, Hybrid-EP, and others. These libraries demonstrate the performance benefits of GPU-initiated RDMA for MoE dispatch and combine operations. This paper presents NCCL EP (Expert Parallelism), a ground-up MoE communication library built entirely on NCCL's Device API. NCCL EP provides unified ncclEpDispatch and ncclEpCombine primitives with both C and Python interfaces, supporting Low-Latency (LL) mode for inference decoding and High-Throughput (HT) mode for training and inference prefill. LL targets small batch sizes (1-128 tokens) using direct all-to-all RDMA+NVLink mesh connectivity with double-buffered communication for overlapping dispatch and combine phases. HT targets large batches (4096+ tokens) using hierarchical communication that aggregates tokens within NVLink domains before inter-node RDMA transmission. Both modes leverage Device API for both intra- and inter-node communications, taking advantage of its topology awareness and optimized GPU-initiated implementation. We evaluate NCCL EP on an H100-based cluster across multi-node configurations, demonstrating competitive LL kernel performance and presenting end-to-end results with vLLM integration. By building MoE communication natively within NCCL, NCCL EP provides a supported path for expert parallelism on current and emerging NVIDIA platforms.

arXiv:2603.28943v1 Announce Type: cross Abstract: This paper presents a hybrid CPU-GPU framework for solving combinatorial scheduling problems formulated as Integer Linear Programming (ILP). While scheduling underpins many optimization tasks in computing systems, solving these problems optimally at scale remains a long-standing challenge due to their NP-hard nature. We introduce a novel approach that combines differentiable optimization with classical ILP solving. Specifically, we utilize differentiable presolving to rapidly generate high-quality partial solutions, which serve as warm-starts for commercial ILP solvers (CPLEX, Gurobi) and rising open-source solver HiGHS. This method enables significantly improved early pruning compared to state-of-the-art standalone solvers. Empirical results across industry-scale benchmarks demonstrate up to a $10\times$ performance gain over baselines, narrowing the optimality gap to $

arXiv:2603.29375v1 Announce Type: cross Abstract: Spacecraft anomaly detection is critical for mission safety, yet deploying sophisticated models on-board presents significant challenges due to hardware constraints. This paper investigates three approaches for spacecraft telemetry anomaly detection -- forecasting & threshold, direct classification, and image classification -- and optimizes them for edge deployment using multi-objective neural architecture optimization on the European Space Agency Anomaly Dataset. Our baseline experiments demonstrate that forecasting & threshold achieves superior detection performance (92.7% Corrected Event-wise F0.5-score (CEF0.5)) [1] compared to alternatives. Through Pareto-optimal architecture optimization, we dramatically reduced computational requirements while maintaining capabilities -- the optimized forecasting & threshold model preserved 88.8% CEF0.5 while reducing RAM usage by 97.1% to just 59 KB and operations by 99.4%. Analysis of deployment viability shows our optimized models require just 0.36-6.25% of CubeSat RAM, making on-board anomaly detection practical even on highly constrained hardware. This research demonstrates that sophisticated anomaly detection capabilities can be successfully deployed within spacecraft edge computing constraints, providing near-instantaneous detection without exceeding hardware limitations or compromising mission safety.

arXiv:2603.29913v1 Announce Type: cross Abstract: The currently dominant AI/ML workloads, such as Large Language Models (LLMs), rely on the efficient execution of General Matrix-Matrix Multiplication (GEMM) operations. Thus, most systems are equipped with dedicated matrix hardware accelerators based on square Systolic Arrays (SAs) of Processing Elements (PEs). While this organization was effective for traditional Deep Neural Networks (DNNs), LLMs introduce input-dependent and highly skewed matrices, leading to underutilized SA resources. To address this challenge, we propose SISA (Scale-In Systolic Array), a novel SA architecture that partitions the traditional square array into horizontal rectangular slabs. With minimal overhead, SISA exposes parallelism through independently scheduled slabs for efficient execution of small or skewed matrix shapes, while retaining full-array operation for large GEMMs. SISA achieves up to 8.52x speedup and 93% energy-delay-product (EDP) reduction for representative LLMs compared to a state-of-the-art monolithic SA with the same number of PEs.

arXiv:2508.00017v4 Announce Type: replace-cross Abstract: We present Generative Logic (GL), a deterministic architecture that starts from user-supplied axiomatic definitions written in a minimalist Mathematical Programming Language (MPL) and systematically explores a configurable region of their deductive neighborhood. Definitions are compiled into a distributed grid of Logic Blocks (LBs) that communicate via a unified hash-based inference engine; whenever the premises of a rule unify, a new fact is emitted with full provenance, yielding replayable, auditable proof graphs. The pipeline includes an Incubator that auto-generates ground-level fact tables, a Compressor that eliminates post-proof redundancy, and an independent external Verifier (34,320 checks, zero failures). Experimental validation on Elementary Number Theory develops Peano arithmetic from axioms and autonomously derives Gauss's summation formula. On commodity hardware, the core proving pipeline completes in under one minute; the full run including Incubator fact generation finishes in approximately ten minutes. The Incubator output further reveals that GL can perform concrete numerical calculations -- each result a proved theorem with full provenance -- opening a path toward a full-provenance Computer Algebra System (CAS). Generated proofs export as navigable HTML for independent inspection. Code, proof graphs, and reproduction instructions are available at github.com/Generative-Logic/GL (commit 6e5b9a4) and archived at doi:10.5281/zenodo.17206386.

arXiv:2603.10062v2 Announce Type: replace-cross Abstract: As LLM agents evolve into collaborative multi-agent systems, their memory requirements grow rapidly in complexity. This position paper frames multi-agent memory as a computer architecture problem. We distinguish shared and distributed memory paradigms, propose a three-layer memory hierarchy (I/O, cache, and memory), and identify two critical protocol gaps: cache sharing across agents and structured memory access control. We argue that the most pressing open challenge is multi-agent memory consistency. Our architectural framing provides a foundation for building reliable, scalable multi-agent systems.

arXiv:2601.07315v4 Announce Type: replace-cross Abstract: Vision Language Models (VLMs) have demonstrated remarkable potential in multimodal reasoning, yet they inherently suffer from spatial blindness and logical hallucinations when interpreting densely structured engineering content, such as analog circuit schematics. To address these challenges, we propose a Vision Language Model-Optimized Collaborative Agent Design Workflow for Analog Circuit Sizing (VLM-CAD) designed for robust, step-by-step reasoning over multimodal evidence. VLM-CAD bridges the modality gap by integrating a neuro-symbolic structural parsing module, Image2Net, which transforms raw pixels into explicit topological graphs and structured JSON representations to anchor VLM interpretation in deterministic facts. To ensure the reliability required for engineering decisions, we further propose ExTuRBO, an Explainable Trust Region Bayesian Optimization method. ExTuRBO serves as an explainable grounding engine, employing agent-generated semantic seeds to warm-start local searches and utilizing Automatic Relevance Determination to provide quantified evidence for the VLM's decisions. Experimental results on two complex circuit benchmarks demonstrate that VLM-CAD significantly enhances spatial reasoning accuracy and maintains physics-based explainability. VLM-CAD consistently satisfies complex specification requirements while achieving low power consumption, with a total runtime under 66 minutes, marking a significant step toward robust, explainable multimodal reasoning in specialized technical domains.

arXiv:2603.13606v2 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) architectures have become essential for scaling large language models, driving the development of specialized device-initiated communication libraries such as DeepEP, Hybrid-EP, and others. These libraries demonstrate the performance benefits of GPU-initiated RDMA for MoE dispatch and combine operations. This paper presents NCCL EP (Expert Parallelism), a ground-up MoE communication library built entirely on NCCL's Device API. NCCL EP provides unified ncclEpDispatch and ncclEpCombine primitives with both C and Python interfaces, supporting Low-Latency (LL) mode for inference decoding and High-Throughput (HT) mode for training and inference prefill. LL targets small batch sizes (1-128 tokens) using direct all-to-all RDMA+NVLink mesh connectivity with double-buffered communication for overlapping dispatch and combine phases. HT targets large batches (4096+ tokens) using hierarchical communication that aggregates tokens within NVLink domains before inter-node RDMA transmission. Both modes leverage Device API for both intra- and inter-node communications, taking advantage of its topology awareness and optimized GPU-initiated implementation. We evaluate NCCL EP on an H100-based cluster across multi-node configurations, demonstrating competitive LL kernel performance and presenting end-to-end results with vLLM integration. By building MoE communication natively within NCCL, NCCL EP provides a supported path for expert parallelism on current and emerging NVIDIA platforms.

arXiv:2603.12269v1 Announce Type: cross Abstract: Early-exit deep neural networks enable adaptive inference by terminating computation when sufficient confidence is achieved, reducing cost for edge AI accelerators in resource-constrained settings. Existing methods, however, rely on suboptimal exit policies, ignore input difficulty, and optimize thresholds independently. This paper introduces DART (Input-Difficulty-Aware Adaptive Threshold), a framework that overcomes these limitations. DART introduces three key innovations: (1) a lightweight difficulty estimation module that quantifies input complexity with minimal computational overhead, (2) a joint exit policy optimization algorithm based on dynamic programming, and (3) an adaptive coefficient management system. Experiments on diverse DNN benchmarks (AlexNet, ResNet-18, VGG-16) demonstrate that DART achieves up to \textbf{3.3$\times$} speedup, \textbf{5.1$\times$} lower energy, and up to \textbf{42\%} lower average power compared to static networks, while preserving competitive accuracy. Extending DART to Vision Transformers (LeViT) yields power (5.0$\times$) and execution-time (3.6$\times$) gains but also accuracy loss (up to 17 percent), underscoring the need for transformer-specific early-exit mechanisms. We further introduce the Difficulty-Aware Efficiency Score (DAES), a novel multi-objective metric, under which DART achieves up to a 14.8 improvement over baselines, highlighting superior accuracy, efficiency, and robustness trade-offs.

arXiv:2603.12739v1 Announce Type: cross Abstract: Spiking Neural Networks (SNNs) have emerged as a biologically inspired alternative to conventional deep networks, offering event-driven and energy-efficient computation. However, their throughput remains constrained by the serial update of neuron membrane states. While many hardware accelerators and Compute-in-Memory (CIM) architectures efficiently parallelize the synaptic operation (W x I) achieving O(1) complexity for matrix-vector multiplication, the subsequent state update step still requires O(N) time to refresh all neuron membrane potentials. This mismatch makes state update the dominant latency and energy bottleneck in SNN inference. To address this challenge, we propose an SRAM-based CIM for SNN with Linear Decay Leaky Integrate-and-Fire (LD-LIF) Neuron that co-optimizes algorithm and hardware. At the algorithmic level, we replace the conventional exponential membrane decay with a linear decay approximation, converting costly multiplications into simple additions while accuracy drops only around 1%. At the architectural level, we introduce an in-memory parallel update scheme that performs in-place decay directly within the SRAM array, eliminating the need for global sequential updates. Evaluated on benchmark SNN workloads, the proposed method achieves a 1.1 x to 16.7 x reduction of SOP energy consumption, while providing 15.9 x to 69 x more energy efficiency, with negligible accuracy loss relative to original decay models. This work highlights that beyond accelerating the (W x I) computation, optimizing state-update dynamics within CIM architectures is essential for scalable, low-power, and real-time neuromorphic processing.

arXiv:2602.11506v3 Announce Type: replace-cross Abstract: The transition toward localized intelligence through Small Language Models (SLMs) has intensified the need for rigorous performance characterization on resource-constrained edge hardware. However, objectively measuring the theoretical performance ceilings of diverse architectures across heterogeneous platforms remains a formidable challenge. In this work, we propose a systematic framework based on the Roofline model that unifies architectural primitives and hardware constraints through the lens of operational intensity (OI). By defining an inference-potential region, we introduce the Relative Inference Potential as a novel metric to compare efficiency differences between Large Language Models (LLMs) on the same hardware substrate. Extensive empirical analysis across diverse compute tiers reveals that variations in performance and OI are significantly influenced by sequence length. We further identify a critical regression in OI as model depth increases. Additionally, our findings highlight an efficiency trap induced by hardware heterogeneity and demonstrate how structural refinements, such as Multi-head Latent Attention (M LA), can effectively unlock latent inference potential across various hardware substrates. These insights provide actionable directions for hardware-software co-design to align neural structures with physical constraints in on-device intelligence. The released code is available in the Appendix C.

arXiv:2601.19940v2 Announce Type: replace-cross Abstract: Among hardware accelerators for deep-learning inference, data flow implementations offer low latency and high throughput capabilities. In these architectures, each neuron is mapped to a dedicated hardware unit, making them well-suited for field-programmable gate array (FPGA) implementation. Previous unrolled implementations mostly focus on fully connected networks because of their simplicity, although it is well known that convolutional neural networks (CNNs) require fewer computations for the same accuracy. When observing the data flow in CNNs, pooling layers and convolutional layers with a stride larger than one, the number of data at their output is reduced with respect to their input. This data reduction strongly affects the data rate in a fully parallel implementation, making hardware units heavily underutilized unless it is handled properly. This work addresses this issue by analyzing the data flow of CNNs and presents a novel approach to designing data-rate-aware, continuous-flow CNN architectures. The proposed approach ensures a high hardware utilization close to 100% by interleaving low data rate signals and sharing hardware units, as well as using the right parallelization to achieve the throughput of a fully parallel implementation. The results show that a significant amount of the arithmetic logic can be saved, which allows implementing complex CNNs like MobileNet on a single FPGA with high throughput.

arXiv:2603.03880v1 Announce Type: cross Abstract: Software-hardware co-design is essential for optimizing in-memory computing (IMC) hardware accelerators for neural networks. However, most existing optimization frameworks target a single workload, leading to highly specialized hardware designs that do not generalize well across models and applications. In contrast, practical deployment scenarios require a single IMC platform that can efficiently support multiple neural network workloads. This work presents a joint hardware-workload co-optimization framework based on an optimized evolutionary algorithm for designing generalized IMC accelerator architectures. By explicitly capturing cross-workload trade-offs rather than optimizing for a single model, the proposed approach significantly reduces the performance gap between workload-specific and generalized IMC designs. The framework is evaluated on both RRAM- and SRAM-based IMC architectures, demonstrating strong robustness and adaptability across diverse design scenarios. Compared to baseline methods, the optimized designs achieve energy-delay-area product (EDAP) reductions of up to 76.2% and 95.5% when optimizing across a small set (4 workloads) and a large set (9 workloads), respectively. The source code of the framework is available at https://github.com/OlgaKrestinskaya/JointHardwareWorkloadOptimizationIMC.

arXiv:2603.02271v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models are an emerging class of workloads critical for robotics and embodied AI at the edge. As these models scale, they demonstrate significant capability gains, yet they must be deployed locally to meet the strict latency requirements of real-time applications. This paper characterizes VLA performance on two generations of edge hardware, viz. the Nvidia Jetson Orin and Thor platforms. Using MolmoAct-7B, a state-of-the-art VLA model, we identify a primary execution bottleneck: up to 75% of end-to-end latency is consumed by the memory-bound action-generation phase. Through analytical modeling and simulations, we project the hardware requirements for scaling to 100B parameter models. We also explore the impact of high-bandwidth memory technologies and processing-in-memory (PIM) as promising future pathways in edge systems for embodied AI.

arXiv:2603.03075v1 Announce Type: cross Abstract: Accurate sea ice mapping is essential for safe maritime navigation in polar regions, where rapidly changing ice conditions require timely and reliable information. While Sentinel-1 Synthetic Aperture Radar (SAR) provides high-resolution, all-weather observations of sea ice, conventional ground-based processing is limited by downlink bandwidth, latency, and energy costs associated with transmitting large volumes of raw data. On-board processing, enabled by dedicated inference chips integrated directly within the satellite payload, offers a transformative alternative by generating actionable sea ice products in orbit. In this context, we present TinyIceNet, a compact semantic segmentation network co-designed for on-board Stage of Development (SOD) mapping from dual-polarized Sentinel-1 SAR imagery under strict hardware and power constraints. Trained on the AI4Arctic dataset, TinyIceNet combines SAR-aware architectural simplifications with low-precision quantization to balance accuracy and efficiency. The model is synthesized using High-Level Synthesis and deployed on a Xilinx Zynq UltraScale+ FPGA platform, demonstrating near-real-time inference with significantly reduced energy consumption. Experimental results show that TinyIceNet achieves 75.216% F1 score on SOD segmentation while reducing energy consumption by 2x compared to full-precision GPU baselines, underscoring the potential of chip-level hardware-algorithm co-design for future spaceborne and edge AI systems.

arXiv:2602.18568v1 Announce Type: cross Abstract: Large language model (LLM) inference performance is increasingly bottlenecked by the memory wall. While GPUs continue to scale raw compute throughput, they struggle to deliver scalable performance for memory bandwidth bound workloads. This challenge is amplified by emerging reasoning LLM applications, where long output sequences, low arithmetic intensity, and tight latency constraints demand significantly higher memory bandwidth. As a result, system utilization drops and energy per inference rises, highlighting the need for an optimized system architecture for scalable memory bandwidth. To address these challenges we present the Reasoning Processing Unit (RPU), a chiplet-based architecture designed to address the challenges of the modern memory wall. RPU introduces: (1) A Capacity-Optimized High-Bandwidth Memory (HBM-CO) that trades capacity for lower energy and cost; (2) a scalable chiplet architecture featuring a bandwidth-first power and area provisioning design; and (3) a decoupled microarchitecture that separates memory, compute, and communication pipelines to sustain high bandwidth utilization. Simulation results show that RPU performs up to 45.3x lower latency and 18.6x higher throughput over an H100 system at ISO-TDP on Llama3-405B.

arXiv:2602.19268v1 Announce Type: cross Abstract: This brief presents a runtime-adaptive, performance-enhanced vector engine featuring a low-resource, iterative CORDIC-based MAC unit for edge AI acceleration. The proposed design enables dynamic reconfiguration between approximate and accurate modes, exploiting the latency-accuracy trade-off for a wide range of workloads. Its resource-efficient approach further enables up to 4x throughput improvement within the same hardware resources by leveraging vectorised, time-multiplexed execution and flexible precision scaling. With a time-multiplexed multi-AF block and a lightweight pooling and normalisation unit, the proposed vector engine supports flexible precision (4/8/16-bit) and high MAC density. The ASIC implementation results show that each MAC stage can save up to 33% of time and 21% of power, with a 256-PE configuration that achieves higher compute density (4.83 TOPS/mm2 ) and energy efficiency (11.67 TOPS/W) than previous state-of-the-art work. A detailed hardware-software co-design methodology for object detection and classification tasks on Pynq-Z2 is discussed to assess the proposed architecture, demonstrating a scalable, energy-efficient solution for edge AI applications.

arXiv:2508.00017v3 Announce Type: replace-cross Abstract: We present Generative Logic (GL), a deterministic architecture that starts from user-supplied axiomatic definitions, written in a minimalist Mathematical Programming Language (MPL), and systematically explores a configurable region of their deductive neighborhood. A defining feature of the architecture is its unified hash-based inference engine, which executes both algebraic manipulations and deterministic logical transformations. Definitions are compiled into a distributed grid of simple Logic Blocks (LBs) that exchange messages; whenever the premises of an inference rule unify, a new fact is emitted with full provenance to its sources, yielding replayable, auditable proof graphs. Experimental validation is performed on Elementary Number Theory (ENT) utilizing a batched execution strategy. Starting from foundational axioms and definitions, the system first develops first-order Peano arithmetic, which is subsequently applied to autonomously derive and prove Gauss's summation formula as a main result. To manage combinatorial explosion, GL algorithmically enumerates conjectures and applies normalization, type constraints, and counterexample (CE) filtering. On commodity hardware, an end-to-end run completes in under 7 minutes. Generated proofs export as navigable HTML so that every inference step can be inspected independently. We outline a hardware-software co-design path toward massively parallel realizations and describe future integration with large language models (LLMs) for auto-formalization and conjecture seeding. The Python, C++, and MPL code to reproduce these experiments, along with the full proof graphs in HTML as well as machine-readable text format, are available in the project's GitHub repository at github.com/Generative-Logic/GL commit 1771330 and are permanently archived at doi:10.5281/zenodo.17206386.

arXiv:2602.14699v1 Announce Type: cross Abstract: This paper envisions a quantum database (Qute) that treats quantum computation as a first-class execution option. Unlike prior simulation-based methods that either run quantum algorithms on classical machines or adapt existing databases for quantum simulation, Qute instead (i) compiles an extended form of SQL into gate-efficient quantum circuits, (ii) employs a hybrid optimizer to dynamically select between quantum and classical execution plans, (iii) introduces selective quantum indexing, and (iv) designs fidelity-preserving storage to mitigate current qubit constraints. We also present a three-stage evolution roadmap toward quantum-native database. Finally, by deploying Qute on a real quantum processor (origin_wukong), we show that it outperforms a classical baseline at scale, and we release an open-source prototype at https://github.com/weAIDB/Qute.

arXiv:2601.18702v4 Announce Type: replace-cross Abstract: The prevailing scaling paradigm of Large Language Models (LLMs) rests on a substrate of "Fuzzy" floating-point arithmetic. To mitigate the inherent instability of this approximate foundation, modern architectures have erected a complex scaffolding of structural and numerical heuristics--Complex Residuals, Pre-RMSNorm, Attention Scaling, and Gradient Clipping--consuming significant compute solely to prevent numerical collapse. We propose a paradigm shift to the "Exact". We introduce the Halo Architecture, grounded in the Rational Field (Q) and powered by a custom Exact Inference Unit (EIU). To resolve the exponential bit-width growth of rational arithmetic, Halo employs a Dual-Ring Topology that unifies two complementary control mechanisms: (1) The Micro-Ring (Continuum Maintenance), which strictly bounds memory complexity via Diophantine Approximation; and (2) The Macro-Ring (Symbolic Alignment), which enforces logical consistency via periodic state collapse. This stable dual-ring substrate allows for the "Great Dismantling" of numerical scaffolding, reducing the Transformer block to its "Clean" algebraic form (Tabula Rasa). Furthermore, we verify the "Efficiency Paradox": the elimination of gradient noise (sigma -> 0) allows for Macro-Learning Rates, potentially reducing the Total Time-to-Convergence by orders of magnitude. Halo demonstrates that General Intelligence requires the hybridization of continuous fields and discrete chains under a rigorous mathematical framework.