Search arXiv⌕ Search

arXiv subjects

Deepu John

Publications and source records attributed to Deepu John.

At least 19 recordsLinked to original sources

Intent-Aware Caching for Efficient LLM Serving

Modern LLM serving systems use prefix caching to accelerate multi-turn conversational workloads. They cache the key-value states of previous conversation prefixes and reuse them when subsequent requests share the same prefix, to avoid redundant prefill computation and reduce serving latency. However, achieving a high cache hit rate remains challenging because cache reuse depends on whether a conversation continues and how soon the next turn arrives. Our analysis of real workloads shows that user intent is a useful signal for this behavior, as different intents exhibit different continuation probabilities and inter-turn delays. Based on this key insight, we propose Intent-Aware Caching (IAC), an online eviction policy that classifies requests by intent, learns lightweight reuse statistics from recent serving logs and leverages these statistics to guide an oracle-inspired eviction strategy. Extensive evaluation on real workloads (WildChat and ShareChat) shows that IAC reaches a 34.6% cache hit rate. This substantially outperforms both the previous learning-based approach (18.8%) and traditional methods such as least recently used (LRU, 12.2%), while reducing average serving latency by up to 46.1%.

cs.AI↗

A Framework for Accelerating Transformer Inference on RISC-V for Edge AI

This work presents a framework for accelerating transformer-based language models (LMs) on resource-constrained IoT devices. The framework targets compact LMs: BERT-Tiny (B-Ty), MobileBERT (M-Bt), MiniLM (M-Lm), Electra (E-Lt) and DeBERTa (D-Bt) -- selected for their architectural diversity and use in edge inference scenarios. The proposed flow derives lightweight instruction set extensions tailored to the non-obvious computational patterns of these models. In addition, a custom instruction is introduced to accelerate the address generation stage of batch matrix multiplication, achieving a performance improvement of 15.39--21.74% with modest ASIC overheads of 6.79% in area and 2.33% in power. To further enhance performance without incurring additional processor core hardware cost, an optional compiler-directed loop unrolling strategy is employed, trading increased code size for overall reduced execution time. Evaluation on the Synopsys trv32p3f RISC-V core demonstrates inference speedups of up to 2.19x, and FPGA implementation on the AMD Zynq UltraScale+ ZCU102 shows a 32.94% area overhead at 75 MHz, whereas the ASIC implementation using the TSMC 28 nm library incurs a 21.08% area overhead while operating at 250 MHz.

cs.AR↗

ZTA-Q: an Open-source RISC-V Platform for Accurate Quantized CNN Inference

Low-precision inference is widely adopted in edge AI to reduce computational cost and memory footprint. However, existing open-source accelerator platforms provide limited end-to-end support for CNNs following the standard TensorFlow Lite integer inference scheme. This paper presents ZTA-Q, an open-source RISC-V-based platform that enables accurate deployment of TensorFlow Lite INT8 models. In addition to extending operator support, ZTA-Q provides a configurable post-processing datapath for studying how circuit-level approximations, including reduced multiplier precision, shared shift scaling, and simplified rounding, affect model accuracy. The proposed system is implemented on a Digilent Arty A7-100T FPGA and operates at 83.3 MHz. Evaluations on representative CNN models show that with LUT, register, and DSP overheads of 26.3%, 12.6%, and 150%, respectively, ZTA-Q limits the degradation in both top-1 and top-5 accuracy to within 0.25 percentage points.

cs.AR↗

AdaptViT: Runtime-Adaptive Vision Transformer Deployment on Custom RISC-V

Deploying Vision Transformers (ViTs) on low-power edge devices is challenging due to high computational demands. Conventional pruning frameworks require a separate compiled binary for each sparsity level, increasing storage overhead and limiting runtime adaptability. This paper presents an end-to-end deployment pipeline that transforms pretrained ViTs into a single runtime-configurable binary, enabling dynamic compute-budget switching on embedded CPUs. This is achieved by restructuring generated C kernels with modified loop bounds and binary-mask control logic, allowing execution to switch across discrete sparsity levels via compact external configuration files. Compared to multi-binary deployment, the proposed runtime-adaptive approach reduces on-device storage by up to 4.86x, requiring only 163 MB for ViT-Base instead of nearly 800 MB. To maximize pruning efficiency, we introduce a hardware-aligned block pruning strategy for Multi-Layer Perceptron (MLP) layers. In addition, a custom ISA extension is proposed to exploit input-reuse patterns in linear projection kernels. On a Synopsys TRV32P3FX RISC-V processor, the full system achieves up to 2.8x speedup at 65% MLP and 50% attention-head pruning for ViT-Base. The ISA extension alone provides a 1.56x speedup and 33% lower inference energy, with a 24.7% area overhead in a TSMC 28 nm implementation.

cs.AR↗

EEG-FuseFormer: A Transformer-Driven Feature Fusion Framework for Seizure Onset Prediction

Epilepsy is one of the most common neurological disorders globally, characterized by recurring seizures and significantly impacting the quality of life. Despite advancements in diagnostic techniques, the mitigation of risks faced by epilepsy patients remains challenging due to the unpredictability of seizure events. An accurate forecast of seizure onset helps to reduce risks in epilepsy patients. In this paper, we propose EEG-FuseFormer, a transformer-based feature fusion framework for seizure-onset prediction that combines intermediate features extracted from Convolutional Neural Networks-Long Short-Term Memory (CNN-LSTM) and ResNet-18 networks. The CNN-LSTM architecture captures both spatial and temporal features directly from the raw signal, whereas the ResNet-18 extracts features from the Short-Time Fourier Transform (STFT) representation of the EEG signals. Fusion is carried out using a transformer encoder, and the final prediction is generated using fully connected dense layers. The CHB-MIT dataset was used to validate the proposed model. The results show that the proposed model achieves a mean recall of 98.85% and outperforms most of the state-of-the-art methods. This study evaluates the ability of the proposed feature fusion model to generalize in cross-patient testing scenarios. Fine-tuning pre-trained models on limited target patient data (target adaptation) within the cross-patient validation framework results in higher recall, precision, and F1-score metrics in comparison to the conventional cross-patient validation approach. Finally, the runtime-based computational complexity of the model is assessed across diverse hardware platforms to highlight the performance-complexity trade-off.

cs.LG↗

WISP: Waste- and Interference-Suppressed Distributed Speculative LLM Serving at the Edge via Dynamic Drafting and SLO-Aware Batching

As Large Language Models (LLMs) become increasingly accessible to end users, an ever-growing number of inference requests are initiated from edge devices and computed on centralized GPU clusters. However, the resulting exponential growth in computation workload is placing significant strain on data centers, while edge devices remain largely underutilized, leading to imbalanced workloads and resource inefficiency across the network. Integrating edge devices into the LLM inference process via speculative decoding helps balance the workload between the edge and the cloud, while maintaining lossless prediction accuracy. In this paper, we identify and formalize two critical bottlenecks that limit the efficiency and scalability of distributed speculative LLM serving: Wasted Drafting Time and Verification Interference. To address these challenges, we propose WISP, an efficient and SLO-aware distributed LLM inference system that consists of an intelligent speculation controller, a verification time estimator, and a verification batch scheduler. These components collaboratively enhance drafting efficiency and optimize verification request scheduling on the server. Extensive numerical results show that WISP improves system capacity by up to 2.1x and 4.1x, and increases system goodput by up to 1.94x and 3.7x, compared to centralized serving and SLED, respectively.

cs.DC↗

Polymorph: Energy-Efficient Multi-Label Classification for Video Streams on Embedded Devices

Real-time multi-label video classification on embedded devices is constrained by limited compute and energy budgets. Yet, video streams exhibit structural properties such as label sparsity, temporal continuity, and label co-occurrence that can be leveraged for more efficient inference. We introduce Polymorph, a context-aware framework that activates a minimal set of lightweight Low Rank Adapters (LoRA) per frame. Each adapter specializes in a subset of classes derived from co-occurrence patterns and is implemented as a LoRA weight over a shared backbone. At runtime, Polymorph dynamically selects and composes only the adapters needed to cover the active labels, avoiding full-model switching and weight merging. This modular strategy improves scalability while reducing latency and energy overhead. Polymorph achieves 40% lower energy consumption and improves mAP by 9 points over strong baselines on the TAO dataset. Polymorph is open source at https://github.com/inference-serving/polymorph/.

cs.CV↗

KWT-Tiny: RISC-V Accelerated, Embedded Keyword Spotting Transformer

This paper explores the adaptation of Transformerbased models for edge devices through the quantisation and hardware acceleration of the ARM Keyword Transformer (KWT) model on a RISC-V platform. The model was targeted to run on 64kB RAM in bare-metal C using a custom-developed edge AI library. KWT-1 was retrained to be 369 times smaller, with only a 10% loss in accuracy through reducing output classes from 35 to 2. The retraining and quantisation reduced model size from 2.42 MB to 1.65 kB. The integration of custom RISC-V instructions that accelerated GELU and SoftMax operations enabled a 5x speedup and thus ~5x power reduction in inference, with inference clock cycle counts decreasing from 26 million to 5.5 million clock cycles while incurring a small area overhead of approximately 29%. The results demonstrate a viable method for porting and accelerating Transformer-based models in low-power IoT devices.

cs.AR↗

Bare-Metal RISC-V + NVDLA SoC for Efficient Deep Learning Inference

This paper presents a novel System-on-Chip (SoC) architecture for accelerating complex deep learning models for edge computing applications through a combination of hardware and software optimisations. The hardware architecture tightly couples the open-source NVIDIA Deep Learning Accelerator (NVDLA) to a 32-bit, 4-stage pipelined RISC-V core from Codasip called uRISC_V. To offload the model acceleration in software, our toolflow generates bare-metal application code (in assembly), overcoming complex OS overheads of previous works that have explored similar architectures. This tightly coupled architecture and bare-metal flow leads to improvements in execution speed and storage efficiency, making it suitable for edge computing solutions. We evaluate the architecture on AMD's ZCU102 FPGA board using NVDLA-small configuration and test the flow using LeNet-5, ResNet-18 and ResNet-50 models. Our results show that these models can perform inference in 4.8 ms, 16.2 ms and 1.1 s respectively, at a system clock frequency of 100 MHz.

cs.AR↗

SLED: A Speculative LLM Decoding Framework for Efficient Edge Serving

The growing gap between the increasing complexity of large language models (LLMs) and the limited computational budgets of edge devices poses a key challenge for efficient on-device inference, despite gradual improvements in hardware capabilities. Existing strategies, such as aggressive quantization, pruning, or remote inference, trade accuracy for efficiency or lead to substantial cost burdens. This position paper introduces a new framework that leverages speculative decoding, previously viewed primarily as a decoding acceleration technique for autoregressive generation of LLMs, as a promising approach specifically adapted for edge computing by orchestrating computation across heterogeneous devices. We propose \acronym, a framework that allows lightweight edge devices to draft multiple candidate tokens locally using diverse draft models, while a single, shared edge server verifies the tokens utilizing a more precise target model. To further increase the efficiency of verification, the edge server batch the diverse verification requests from devices. This approach supports device heterogeneity and reduces server-side memory footprint by sharing the same upstream target model across multiple devices. Our initial experiments with Jetson Orin Nano, Raspberry Pi 4B/5, and an edge server equipped with 4 Nvidia A100 GPUs indicate substantial benefits: 2.2 more system throughput, 2.8 more system capacity, and better cost efficiency, all without sacrificing model accuracy.

cs.DC↗

TinyDrop: Tiny Model Guided Token Dropping for Vision Transformers

Vision Transformers (ViTs) achieve strong performance in image classification but incur high computational costs from processing all image tokens. To reduce inference costs in large ViTs without compromising accuracy, we propose TinyDrop, a training-free token dropping framework guided by a lightweight vision model. The guidance model estimates the importance of tokens while performing inference, thereby selectively discarding low-importance tokens if large vit models need to perform attention calculations. The framework operates plug-and-play, requires no architectural modifications, and is compatible with diverse ViT architectures. Evaluations on standard image classification benchmarks demonstrate that our framework reduces FLOPs by up to 80% for ViTs with minimal accuracy degradation, highlighting its generalization capability and practical utility for efficient ViT-based classification.

cs.CV↗

Optimal Brain Connection: Towards Efficient Structural Pruning

Structural pruning has been widely studied for its effectiveness in compressing neural networks. However, existing methods often neglect the interconnections among parameters. To address this limitation, this paper proposes a structural pruning framework termed Optimal Brain Connection. First, we introduce the Jacobian Criterion, a first-order metric for evaluating the saliency of structural parameters. Unlike existing first-order methods that assess parameters in isolation, our criterion explicitly captures both intra-component interactions and inter-layer dependencies. Second, we propose the Equivalent Pruning mechanism, which utilizes autoencoders to retain the contributions of all original connection--including pruned ones--during fine-tuning. Experimental results demonstrate that the Jacobian Criterion outperforms several popular metrics in preserving model performance, while the Equivalent Pruning mechanism effectively mitigates performance degradation after fine-tuning. Code: https://github.com/ShaowuChen/Optimal_Brain_Connection

cs.CV↗

MARVEL: An End-to-End Framework for Generating Model-Class Aware Custom RISC-V Extensions for Lightweight AI

Deploying deep neural networks (DNNs) on resource-constrained IoT devices remains a challenging problem, often requiring hardware modifications tailored to individual AI models. Existing accelerator-generation tools, such as AMD's FINN, do not adequately address extreme resource limitations faced by IoT endpoints operating in bare-metal environments without an operating system (OS). To overcome these constraints, we propose MARVEL-an automated, end-to-end framework that generates custom RISC-V ISA extensions tailored to specific DNN model classes, with a primary focus on convolutional neural networks (CNNs). The proposed method profiles high-level DNN representations in Python and generates an ISA-extended RISC-V core with associated compiler tools for efficient deployment. The flow leverages (1) Apache TVM for translating high-level Python-based DNN models into optimized C code, (2) Synopsys ASIP Designer for identifying compute-intensive kernels, modeling, and generating a custom RISC-V and (3) Xilinx Vivado for FPGA implementation. Beyond a model class specific RISC-V, our approach produces an optimized bare-metal C implementation, eliminating the need for an OS or extensive software dependencies. Unlike conventional deployment pipelines relying on TensorFlow/PyTorch runtimes, our solution enables seamless execution in highly resource-constrained environments. We evaluated the flow on popular DNN models such as LeNet-5*, MobileNetV1, ResNet50, VGG16, MobileNetV2 and DenseNet121 using the Synopsys trv32p3 RISC-V core as a baseline. Results show a 2x speedup in inference and upto 2x reduction in energy per inference at a 28.23% area overhead when implemented on an AMD Zynq UltraScale+ ZCU104 FPGA platform.

cs.AR↗

QPART: Adaptive Model Quantization and Dynamic Workload Balancing for Accuracy-aware Edge Inference

As machine learning inferences increasingly move to edge devices, adapting to diverse computational capabilities, hardware, and memory constraints becomes more critical. Instead of relying on a pre-trained model fixed for all future inference queries across diverse edge devices, we argue that planning an inference pattern with a request-specific model tailored to the device's computational capacity, accuracy requirements, and time constraints is more cost-efficient and robust to diverse scenarios. To this end, we propose an accuracy-aware and workload-balanced inference system that integrates joint model quantization and inference partitioning. In this approach, the server dynamically responds to inference queries by sending a quantized model and adaptively sharing the inference workload with the device. Meanwhile, the device's computational power, channel capacity, and accuracy requirements are considered when deciding. Furthermore, we introduce a new optimization framework for the inference system, incorporating joint model quantization and partitioning. Our approach optimizes layer-wise quantization bit width and partition points to minimize time consumption and cost while accounting for varying accuracy requirements of tasks through an accuracy degradation metric in our optimization model. To our knowledge, this work represents the first exploration of optimizing quantization layer-wise bit-width in the inference serving system, by introducing theoretical measurement of accuracy degradation. Simulation results demonstrate a substantial reduction in overall time and power consumption, with computation payloads decreasing by over 80% and accuracy degradation kept below 1%.

cs.DC↗

ORXE: Orchestrating Experts for Dynamically Configurable Efficiency

This paper presents ORXE, a modular and adaptable framework for achieving real-time configurable efficiency in AI models. By leveraging a collection of pre-trained experts with diverse computational costs and performance levels, ORXE dynamically adjusts inference pathways based on the complexity of input samples. Unlike conventional approaches that require complex metamodel training, ORXE achieves high efficiency and flexibility without complicating the development process. The proposed system utilizes a confidence-based gating mechanism to allocate appropriate computational resources for each input. ORXE also supports adjustments to the preference between inference cost and prediction performance across a wide range during runtime. We implemented a training-free ORXE system for image classification tasks, evaluating its efficiency and accuracy across various devices. The results demonstrate that ORXE achieves superior performance compared to individual experts and other dynamic models in most cases. This approach can be extended to other applications, providing a scalable solution for diverse real-world deployment scenarios.

cs.CV↗

DyCE: Dynamically Configurable Exiting for Deep Learning Compression and Real-time Scaling

Conventional deep learning (DL) model compression and scaling methods focus on altering the model's components, impacting the results across all samples uniformly. However, since samples vary in difficulty, a dynamic model that adapts computation based on sample complexity offers a novel perspective for compression and scaling. Despite this potential, existing dynamic models are typically monolithic and model-specific, limiting their generalizability as broad compression and scaling methods. Additionally, most deployed DL systems are fixed, unable to adjust their scale once deployed and, therefore, cannot adapt to the varying real-time demands. This paper introduces DyCE, a dynamically configurable system that can adjust the performance-complexity trade-off of a DL model at runtime without requiring re-initialization or redeployment on inference hardware. DyCE achieves this by adding small exit networks to intermediate layers of the original model, allowing computation to terminate early if acceptable results are obtained. DyCE also decouples the design of an efficient dynamic model, facilitating easy adaptation to new base models and potential general use in compression and scaling. We also propose methods for generating optimized configurations and determining the types and positions of exit networks to achieve desired performance and complexity trade-offs. By enabling simple configuration switching, DyCE provides fine-grained performance tuning in real-time. We demonstrate the effectiveness of DyCE through image classification tasks using deep convolutional neural networks (CNNs). DyCE significantly reduces computational complexity by 23.5% for ResNet152 and 25.9% for ConvNextv2-tiny on ImageNet, with accuracy reductions of less than 0.5%.

cs.LG↗

DCentNet: Decentralized Multistage Biomedical Signal Classification using Early Exits

DCentNet is a novel decentralized multistage signal classification approach designed for biomedical data from IoT wearable sensors, integrating early exit points (EEP) to enhance energy efficiency and processing speed. Unlike traditional centralized processing methods, which result in high energy consumption and latency, DCentNet partitions a single CNN model into multiple sub-networks using EEPs. By introducing encoder-decoder pairs at EEPs, the system compresses large feature maps before transmission, significantly reducing wireless data transfer and power usage. If an input is confidently classified at an EEP, processing stops early, optimizing efficiency. Initial sub-networks can be deployed on fog or edge devices to further minimize energy consumption. A genetic algorithm is used to optimize EEP placement, balancing performance and complexity. Experimental results on ECG classification show that with one EEP, DCentNet reduces wireless data transmission by 94.54% and complexity by 21%, while maintaining original accuracy and sensitivity. With two EEPs, sensitivity reaches 98.36%, accuracy 97.74%, wireless data transmission decreases by 91.86%, and complexity is reduced by 22%. Implemented on an ARM Cortex-M4 MCU, DCentNet achieves an average power saving of 73.6% compared to continuous wireless ECG transmission.

eess.SP↗

Tiny Models are the Computational Saver for Large Models

This paper introduces TinySaver, an early-exit-like dynamic model compression approach which employs tiny models to substitute large models adaptively. Distinct from traditional compression techniques, dynamic methods like TinySaver can leverage the difficulty differences to allow certain inputs to complete their inference processes early, thereby conserving computational resources. Most existing early exit designs are implemented by attaching additional network branches to the model's backbone. Our study, however, reveals that completely independent tiny models can replace a substantial portion of the larger models' job with minimal impact on performance. Employing them as the first exit can remarkably enhance computational efficiency. By searching and employing the most appropriate tiny model as the computational saver for a given large model, the proposed approaches work as a novel and generic method to model compression. This finding will help the research community in exploring new compression methods to address the escalating computational demands posed by rapidly evolving AI models. Our evaluation of this approach in ImageNet-1k classification demonstrates its potential to reduce the number of compute operations by up to 90\%, with only negligible losses in performance, across various modern vision models.

cs.AI↗