Compute-in-Memory

Analog in-memory computing is a promising platform for on-device execution of large language models because it performs matrix--vector multiplications (MVMs) in memory and in parallel, reducing data movement. However, limited digital-to-analog converter precision, input noise, and finite conductance states can degrade model accuracy, while full-model retraining to address these effects can be costly. We develop an optimizer-agnostic, parameter-efficient adaptation method based on Low-Rank Adaptation (LoRA), keeping the pretrained weights stored on analog arrays fixed while training the LoRA weights to adapt to downstream tasks and hardware non-idealities. Reliable adaptation requires handling errors in both forward and backward MVMs and physical weight updates. We use input reshaping to reduce input-induced MVM errors and update accumulation to retain small updates before programming them to finite-state analog devices. Across Llama-3.2-1B-Instruct and Llama-3-8B with both Muon and AdamW, input reshaping improves analog LoRA fine-tuning under noisy MVM computation. Update accumulation separately preserves sub-threshold updates and substantially improves adaptation under finite-resolution programming, including configurations with as few as 20 conductance states. Additional experiments show consistent held-out negative log-likelihood improvements across noisy analog settings.

Analog in-memory computing (IMC) offers a promising path toward energy-efficient large language model (LLM) inference by executing matrix multiplications (MatMul) directly within memory arrays in the analog domain. Its efficiency, however, comes with an additional source of error: limited-precision analog-to-digital converters (ADCs) quantize accumulated analog partial sums, introducing output-side error distinct from conventional activation and weight quantization at the MatMul inputs. Clipping can mitigate both operand and ADC quantization errors, but the optimal clipping factors must jointly balance activation rounding and clipping, weight rounding and clipping, and ADC quantization. Existing clipping methods, designed for digital quantization, do not explicitly optimize these coupled sources of IMC error and often rely on costly search-based calibration. We introduce IMC-CLINIC (Coupled Loss-Informed Newton Iterations for Clipping), a clipping calibration framework based on an analytical surrogate for IMC MatMul output error. The surrogate jointly models operand quantization, accumulated clipping-induced bias, and ADC quantization, enabling efficient evaluation of its gradient and approximate curvature from a small calibration set. IMC-CLINIC jointly optimizes activation and weight clipping factors using a safeguarded Newton-type method. Across multiple models and datasets, it improves average zero-shot accuracy by 6.5-11.5 percentage points over the grid search baseline while reducing calibration time by factors of 10.0-12.1. Its analytical surrogate closely tracks empirical IMC output error, and its optimizer is certified within 1% of the global optimum under the loss objective across all projections on two representative models.

Large Language Models (LLMs) have shown impressive capabilities across a range of natural language processing tasks, and LLM inference has emerged as a critical workload for enabling downstream applications. The demands of serving LLM inference are becoming increasingly challenging as requests shift toward longer sequences and heavier inference, driven by retrieval-augmented generation, inference-time compute scaling, and long-context applications. Additionally, these challenges are compounded by hardware trends, as memory capacity and communication bandwidth are not scaling as fast as increases in workload complexity. Compute-in-Flash is a promising solution to address memory bandwidth limitations by moving computation close to memory, and to exploit the large capacity of SSD technologies. However, it is challenging to deploy LLMs on these systems as they lack support for high-precision floating point operations and have limited write endurance. In our work, we aim to address these challenges by designing inference algorithms to enable LLM inference on Flash compute-in-memory devices. We present an end-to-end integer-only quantization approach to eliminate expensive floating-point computations. To address the limited write endurance, we design a dictionary-based KV cache compression strategy based on sparse dictionary coding that represents each KV vector as a linear combination of static dictionary vectors. These algorithmic improvements enable us to exploit the benefits of Compute-in-Flash for both model weights and KV cache, and to minimize expensive data transfer operations. Across Llama-3.1-8B and Qwen-2.5-7B, our combined method exhibits limited accuracy degradation while reducing dynamic KV cache traffic by 15$\times$.

Learning and decision-making in animals are often modeled as Bayesian processes, where sensory evidence is integrated with prior beliefs to guide behavior in the face of uncertainty. But what are the inherent neural dynamics that give rise to this ability, and how could they be replicated in computing systems? This abstract discusses a biologically grounded framework in which noisy neural and synaptic dynamics perform inference and learning via stochastic sampling from an internal energy function, capturing uncertainty over latent states and model parameters through neural and synaptic variability, respectively. This enables approaches such as predictive coding networks to account for epistemic uncertainty via Markov chain Monte Carlo sampling. Drawing a parallel between intrinsic noise in biological systems and electrical noise in emerging probabilistic analogue memory technologies, we highlight how analogue in-memory computing hardware naturally emerges as the solution for massively scalable and energy-efficient probabilistic inference.

Artificial neural networks rely on vector-matrix multiplications (VMMs), whose implementation in von Neumann architectures is dominated by costly data movement between memory and processing units. Spiking neural networks (SNNs) mitigate this bottleneck by performing in-memory, analog VMMs using memristive crossbar arrays. However, conventional current-mode readout circuits incur significant area and power overhead. This work proposes a fully analog readout architecture based on voltage-to-time conversion of the VMM output. By sensing the column voltage, the proposed approach avoids current-mode summing and scaling circuitry, improving area and energy efficiency. Post-layout simulations of a 10x1 SNN implemented in a 130 nm CMOS technology validate the proposed architecture, while application to a trained 64x10 SNN for digit classification further demonstrates its feasibility for SNN inference.

Analog in-memory computing (AIMC) speeds up neural-network inference by doing the arithmetic directly inside a memory array, instead of shuttling weights back and forth between memory and a processor. This saves energy, but the physical devices that store the weights are imperfect: programming errors, electrical noise, limited-resolution converters, and outright broken cells all distort the computation, and every physical chip is distorted in its own way. A designer with several such chips available faces an uncomfortable choice: run all of them and combine the answers (safe, but wasteful of energy), or trust a single chip blindly (cheap, but with no guarantee on how often it is wrong). This paper introduces RACE-AIMC (Risk-Aware Certified Ensemble for AIMC), a framework that resolves this choice with statistics rather than guesswork. Offline, RACE-AIMC studies a pool of physical accelerators, picks the single best one for a given energy budget, and computes a mathematically exact upper bound on how often that accelerator will be wrong when it chooses to answer. Online, only that one accelerator is switched on; a lightweight check decides whether to accept its answer or defer to a fallback. In our simulations using a noisy weight mapping and multiple independent test runs, every certified bound stayed under a 10% error target (mean bound 7.83% +- 0.89%, with 70.88% +- 0.98% of inputs answered directly). The resulting system matches the accuracy of a clean digital baseline while cutting modeled energy use by 69.02% relative to always running every accelerator in the pool.

Diffusion models achieve strong image generation quality but incur high iterative denoising costs. Analog compute-in-memory (CIM) can accelerate matrix-vector multiplications, yet spatial memory variations perturb weights and accumulate during sampling. Unlike conventional neural networks, diffusion models' temporal sensitivity to hardware noise remains underexplored. We investigate diffusion inference using a noise model calibrated and validated against measurements collected from multiple physical CIM chips. Our results show that the early, high-noise denoising stage is substantially more vulnerable than the final refinement stage. A first-order trajectory analysis attributes this behavior to the repeated propagation of correlated prediction errors induced by a fixed hardware mapping. Based on this observation, we propose ASSERT, a training-free sampler that uses higher stochasticity early and smoothly transitions to deterministic denoising. The injected stochasticity changes subsequent activation trajectories and thereby reduces their alignment with persistent spatial errors. Across the evaluated settings, ASSERT achieves up to 2.58$\times$ lower FID than deterministic DDIM on high-resolution datasets and 7.68$\times$ lower FID in the CIFAR-10 step-count study, without changing model parameters or the number of network evaluations.

Analog compute-in-memory (CIM) enables energy-efficient neural network inference, but device variation and read noise can severely degrade low-bit quantized models. Existing CIM-oriented quantization methods mainly minimize ideal quantization error, ignoring the hardware noise floor and thus causing inefficient precision allocation. We propose NANQ, a noise-aware mixed-precision non-uniform quantization framework for analog CIM. NANQ models magnitude-dependent weight noise from measured responses of an eFlash CIM array and converts the noise profile into an adaptive quantization density, assigning finer resolution to low-noise regions while avoiding ineffective precision in noise-dominated regions. It further assigns layer-wise bit-widths by identifying each layer's precision saturation point under hardware noise using a unified threshold. On-chip experiments on an eFlash CIM SoC show that, under 2-bit weight-magnitude quantization, NANQ improves vision-model accuracy by 8.05 percentage points and reduces language-model PPL by 54.7% on average over PowerQuant. Mixed-precision NANQ captures most of the gains obtainable from additional quantization resources with only 3.2-3.8 equivalent bits.

We show with experiments and system-level simulations that it is possible to successfully mitigate the impact of retention loss on inference accuracy degradation by using both circuit-level compensation techniques and batch normalization recalibration at the algorithmic level. Experiments are performed on a single-poly floating-gate (FG) analog non-volatile memory array for analog in-memory computing fabricated in a standard 65 nm CMOS. We use a model of retention-loss statistics calibrated with experiments to evaluate the system-level impact on neural network models such as VGG-10/CIFAR-10 and WideResNet-28-10/CIFAR-100. We show that, after 60 days since programming, combined mitigation techniques enable to recover the baseline inference accuracy within 2-4%

Monte Carlo tree search (MCTS) enables artificial intelligence (AI) decision-making, but requires 55-300 W on conventional processors, limiting edge deployment. In-memory computing (IMC) is energy-efficient on regular workloads but has been considered incompatible with irregular multi-phase algorithms. We introduce phase-to-primitive decomposition, which reformulates each algorithmic phase as a hardware-native IMC primitive. Applied to MCTS, selection, expansion, rollout and backpropagation map to content-addressable memory, combinational logic, a resistive random-access memory (RRAM) crossbar and static random-access memory, keeping search on chip. At 22 nm with fabricated RRAM-array parameters, IMC-MCTS consumes ~60 mW for 9x9 Go, achieving 96x energy efficiency over a central processing unit (CPU) and 65x-2,059x over an H100 graphics processing unit (GPU). It reaches a European Go Federation rating within sample-size uncertainty of open-source Go engines (Pachi-UCT and Michi-C). The same substrate runs eight applications across four AI domains.

Transformers are widely used across many domains, including natural language processing, computer vision, web search, and DNA sequence analysis. Given their broad applicability, improving the performance of transformer models is critical. However, the high volume of data movement between processing units and memory during attention operations significantly limits their efficiency. Processing-In-Memory (PIM) mitigates this issue by performing computations directly inside memory. While prior work has proposed PIM-based transformer implementations, they suffer from costly inter-bank communication, and struggle to scale due to the limited capacity of memory banks. As a result, attention-related data must be split across banks, diminishing the potential benefits of PIM. In this work, we propose RED-PIM, an algorithm-architecture co-design that reduces attention latency by minimizing inter-bank data movement from O(N^2) to O(N) and shrinking intermediate attention matrices from N x N to d x d. By reorganizing matrix operations, performing computations locally, and employing an optimized data transfer strategy, RED-PIM significantly reduces computation cost and interconnect traffic. Compared to baseline PIM implementation, RED-PIM achieves inference time reductions ranging from 16.05% to 99.99% (geometric mean of 66.42%), with the largest gains on longer sequences. On real-world datasets, RED-PIM improves performance by 99.60% for long documents and 13.44% for shorter ones, while maintaining or improving accuracy. These results demonstrate RED-PIM's effectiveness for scalable and efficient transformer inference.

Several edge computing platforms, such as autonomous vehicles and smart sensing devices, need to adapt to dynamic environments in real time by learning from new data in the field. Continual learning has emerged as a promising solution for edge training, by incorporating techniques that successfully combine a highly summarized version of previously trained data (to avoid catastrophic forgetting) with recently sensed data. However, as is the case with other ML algorithms, continual learning generates significant data movement between general-purpose CPUs/GPUs and memory, impacting the suitability of continual learning for edge platforms. In-memory computing (IMC; also known as processing-using-memory) can curtail this waste and make continual learning feasible at the edge, but it faces two unique challenges: (1) IMC architectures make use of noisy computation operations that significantly harm training accuracy; and (2) IMC architectures have poor and often incomplete support for resource-efficient training. To address these challenges, we propose CLASP (the Continual Learning Acceleration System Platform), which to our knowledge is the first end-to-end system with IMC acceleration for continual learning. The hardware and software of CLASP are co-designed to support a wide range of continual learning algorithms, through software-visible assembly-level instructions that can be incorporated without constraints into ML-based algorithms. CLASP is designed around a back-end-of-line (BEOL) compatible ECRAM device that we fabricate, which can overcome the challenges of IMC-based training using other emerging memory devices. We show that CLASP with ECRAM approaches the accuracy of in-GPU training, while delivering a speedup of 67x and energy savings of 132x for learning without forgetting and experience replay using MNIST.

Processing-In-Memory (PIM) has emerged as a promising technology for accelerating machine learning (ML) workloads. Specifically, non-volatile memory-based PIM architectures have enabled effective ML acceleration due to their ability to perform energy-efficient matrix-vector multiplication operations. However, these devices suffer from non-idealities such as thermal noise. This noise alters the stored values in the memory cells which correspond to actual model weights, compromising the inference accuracy. In this work, we introduce ThRIve, a noise-aware training methodology that leverages low-rank adaptation to enable thermally robust inference on heterogeneous PIM architectures. ThRIve selectively stores these low-rank noise-aware parameters on a hardware that is less susceptible to thermal noise, enabling robustness against temperature-induced noise variations. ThRIve mitigates the effects of thermal-noise and prevent the drop in inference accuracy across the entire operating temperature range. Experimental results demonstrate that ThRIve-enabled architectures maintain consistent inference accuracy, with the mean accuracy staying within 2% of the ideal (i.e., noise-free) accuracy, and the variation in accuracy across the entire operating temperature range remaining within 2% of the mean. The proposed methodology achieves accuracy and robustness comparable to thermally-resilient Static Random-Access Memory (SRAM)-based PIM systems, while delivering up to 5.4x reduction in energy-delay product (EDP) during CNN model inferencing.

Recent FPGAs have improved deep learning (DL) inference efficiency through dedicated tensor blocks and in-BRAM computation. ReRAM-based analog in-memory computing (IMC) pushes efficiency further, offering an order-of-magnitude improvement in compute density and energy efficiency over conventional digital logic by performing vector-matrix multiplication (VMM) directly within the ReRAM crossbar; prior work has integrated such IMC blocks into FPGAs for DL inference. However, conventional IMC designs support only static-weight VMM, leaving nonlinear operations and dynamic matrix-matrix multiplication (DIMM) to the FPGA fabric. As a result, the benefits of IMC are largely confined to static-weight models, whereas Transformer-based models, which rely on frequent nonlinear and DIMM operations, gain only limited improvement. Moreover, the ADCs within each IMC block consume more than 70% of its area and power, further limiting system efficiency and scalability. To address these limitations, we propose a novel FPGA architecture that integrates an ADC-free IMC block, replacing the conventional ADC with analog content-addressable memories (ACAMs) that natively perform nonlinear operations inside the block. To fully exploit this block, we conduct an FPGA-aware design-space exploration that determines optimal crossbar dimensions while balancing FPGA area, flexibility, and DL performance, and we develop an efficient mapping that leverages ACAMs to carry out DIMM operations, extending the applicability of IMC to attention computation. On CNN and Transformer-based benchmarks, the proposed architecture achieves up to 40x and 1.9x higher energy efficiency and 4.1x and 2.5x higher area efficiency, respectively. Overall, it significantly improves FPGA DL inference efficiency and sustains robust gains on Transformer-based workloads across long input sequences, advancing domain-specialized FPGA design.

Computing-in-Memory (CIM) accelerators execute Matrix-Vector Multiplications (MVMs) in memory, making them a compelling solution for Machine Learning (ML) workloads. However, existing ML workload partitioning approaches for CIM accelerators do not fully account for Resistive Random Access Memory (RRAM) constraints such as limited memory, high write latency, and limited endurance. They also neglect parallelism, low-level architectural effects, or the Central Processing Unit (CPU) as a complementary compute resource. To address these limitations, we propose an Integer Linear Programming (ILP)-based workload partitioning framework for heterogeneous CPU-CIM systems. It minimizes end-to-end inference latency under RRAM constraints, captures parallelism, and combines empirical profiling with analytical models. Using our framework, heterogeneous CPU-CIM execution achieves speedups of up to 30.9x over CPU-only execution on an edge CPU and 7.3x over a high-performance CPU. A Design Space Exploration (DSE) yields further design insights for future CIM accelerators.

Building on resistive communication, this paper presents a physics-based design of an on-chip neural network with multi-level memristive synapses supporting a dense spectrum of conductance states. Derived from ionic transport physics, we develop a state-variable model and quantify storable sub-levels under thermal noise, drift, and quantized conductance. We assemble these devices into a 1T1R crossbar fabric, derive the linear algebra of analog vector-matrix multiplication (VMM) under wire resistance, and design a differential synapse for signed weights. A multilayer pipeline executes inference, backpropagation, and weight updates physically in the analog domain. We derive the in-situ outer-product learning rule, its discretization onto the conductance lattice, and the resulting quantization noise. We provide energy, area, capacity, and inter-tile models, showing this substrate is ideally suited for large language models (LLMs). Our design eliminates weight movement, surpassing binary ReRAM and traditional CMOS. We detail the material stack (HfO_2-based), the FEOL/BEOL CMOS foundry-integration flow, a self-contained SPICE model, the complete memristive-FPGA neuromorphic system, and an in-memory self-attention engine with current-mode translinear softmax. Finally, a ternary BitNet datapath shows projected per-token efficiency orders of magnitude better than advanced CPUs or GPUs. The result is a self-contained hardware-native blueprint for a high-density, analog, in-memory neural processor.

Traditional CPU, GPU, and NPU architectures are increasingly limited by the von Neumann bottleneck. While In-Memory Computing (IMC) using ReRAM crossbar arrays offers a high-density, energy-efficient alternative, its practical deployment is constrained through their non-idealities. Existing hardware-aware training frameworks often require training from scratch, which is computationally prohibitive for modern large-scale models. In this work, we propose a finetuning-based hardware-aware training algorithm that enables robust DNN deployment on ReRAM with minimal training overhead. Our approach mitigates I-V non-linearity by applying a range-shrunk sinh transformation and incorporates retention errors directly into a regularization loss during the finetuning process. We evaluate our framework across models and tasks such as image classification and question-answering (QA). Experimental results demonstrate that our method achieves similar accuracy on large-scale models like ResNet18 and DeiT-Tiny as the base model. In-case of ImageNet for MobileNetV3 families the technique has only less than 2% accuracy degradation. Further, applying the technique on the SQuAD v2 dataset results in only 1 point degradation of F-1 score.

Neural dynamical systems are expressive temporal predictors that capture continuous-time dynamics through fine-grained state updates. However, this sequential structure maps poorly onto digital hardware optimized for dense matrix operations, a mismatch that analog neuromorphic computing, with its native continuous-time dynamics, can resolve. We introduce FerroNDS, a neuromorphic system built from two analog primitives: an integrator for temporal accumulation and an oscillator for frequency-selective filtering. We map this system onto compute-in-memory hardware based on multi-bit ferrodiodes. A 128-neuron instance of FerroNDS computes short-time Fourier transform and forecasts a 500-ms horizon for periodic, quasi-periodic, and chaotic signals. The system achieves sub-watt real-time operation with per-neuron per-inference energy of 1.64 $μ$J (200 Hz) and 0.29 $μ$J (10 kHz), 25-40$\times$ area reduction over SRAM-based digital systems, and per-layer latency of 3.18 ms (200 Hz) and 63.87 $μ$s (10 kHz). To our knowledge, this is the first end-to-end integration of a ferrodiode into a neuromorphic computational framework, establishing ferroelectric compute-in-memory as a practical substrate for analog neural dynamical systems.

The deployment of modern machine learning (ML) solutions on resource-constrained edge devices highlights implementation challenges. This is especially true for extreme edge applications that include safety-critical components, such as autonomous navigation tasks. This paper demonstrates an artificial neural network (ANN) design leveraging Metal-Oxide Resistive RAM (RRAM) -based Analogue Content Addressable Memory (ACAM) as an efficient hardware substrate for performing metric-based classification and online adaptation on the edge. The proposed design is based on a custom Template piXeL (TXL) cell used for building the ACAM module, where each TXL cell acts as a configurable receptive field neuron. These cells employ a Radial Basis activation function to calculate the distance of an input from the programmed receptive field. The TXL can be organised into dense arrays for calculating the distance of a high-dimensional input against all stored prototypes, effectively performing fast and energy efficient similarity search. This hardware engine enables on-the-fly learning, where the receptive field parameters can be tuned to track domain shift. Through simulation of the proposed TXL-RBF classifier we can achieve 89.1% accuracy on the MNIST dataset while consuming 185fJ per cell per operation when operating at 10MHz.

Deep neural networks (DNNs) have achieved state-of-the-art performance across diverse domains. However, typical Von Neumann compute paradigms face severe memory bottlenecks. Emerging near-memory and compute-in-memory approaches alleviate this but incur significant peripheral overhead. Computational Random Access Memory (CRAM) based on MRAM enables in-situ logic without peripheral overhead, offering a dense, energy-efficient solution. However, probabilistic MRAM switching induces gate-level errors that limit the scalability and reliability of CRAM for accelerating DNN. Moreover, the large number of sequential MRAM writes severely constrains CRAM throughput. To address these challenges, we propose an error-resilient CRAM (CRAM-ER) architecture for scalable in-memory matrix-vector multiplications (MVMs). Our error-aware hardware-software co-design framework leverages a hybrid spintronic-CRAM + CMOS adder-tree architecture to mitigate the impact of device-level errors, demonstrating MVM functionality with high area and energy efficiency. We further develop an error-aware model fine-tuning and fine-grained error correction for enhanced error resilience. Evaluations of the CMOS+spintronic hybrid architecture on DNN benchmarks show near-lossless accuracy while reducing CRAM latency by up to 2 orders of magnitude, outperforming CPU/GPU+high-bandwidth DRAM in both energy efficiency and energy-delay product.

This tech note describes the architecture and execution results of the LPDDR5X-PIM simulator, developed by Samsung Electronics. Based on the latest research and internal specifications, the simulator provides a high-fidelity model of both the hardware data paths and the software control layers of the LPDDR5X-PIM block. This integrated hardware-software simulation approach enables precise evaluation of system performance and energy efficiency while maximizing PIM resource utilization. We have refined existing simulation frameworks to align with actual hardware implementation, ensuring consistent behavioral accuracy. Further technical details regarding the specific architecture and circuit design of the LPDDR5X-PIM will be disclosed in future publications

Spiking neural networks (SNNs) provide event-driven and low-power computation inspired by biological neural systems, but current implementations rely on von Neumann graphics processing units (GPUs) and central processing units (CPUs) platforms, where memory and computation bottlenecks limit energy efficiency. To address this challenge, this paper proposes an analog memristor-based spiking neural network (SNN) accelerator that integrates in-memory synaptic computation with analog integrate-and-fire (IF) neurons, eliminating multi-transistor CMOS synapse circuits and enabling asynchronous event-driven operation at the 45nm technology node. Additionally, a digital SNN accelerator is designed and optimized at the 5 nm technology node for comparison. The proposed architecture is evaluated using a predator-prey tracking task that emulates pursuit behavior. In this task, the analog SNN accelerator's inference closely matches the ideal software inference with a mean squared error (MSE) of 0.004. HSPICE simulation results show that the proposed analog SNN accelerator achieves 12.7 times lower energy consumption and 1.26 times lower delay compared to the digital baseline, demonstrating the potential of memristor-based neuromorphic circuits for energy-efficient real-time edge intelligence.

This work presents E-ReCON, a 16 Kb energy and resource-efficient digital compute-in-memory (DCIM) macro based on a compact 3T1R ReRAM bitcell for edge-AI inference. The proposed bitcell occupies only 0.85 um^2 and supports reliable AND-based in-memory multiplication for both conventional convolutional neural network (CNN) and spiking neural network (SNN) workloads. To reduce accumulation overhead, a novel interleaved 10T/28T adder tree is introduced, reducing transistor count and power consumption by 37% and 28%, respectively, compared to a conventional 28T RCA-based design. Implemented in 65 nm CMOS at 1.2 V, the proposed macro achieves a minimum latency of 0.48 ns, throughput of 2.31-3.1 TOPS, and energy efficiency of up to 419 TOPS/W. When evaluated on LeNet-5, AlexNet, and CNN-8 models, the macro achieves 97.81%, 93.23%, and 96.51% accuracy on MNIST/A-Z, CIFAR10, and SVHN datasets, respectively. In addition, 40% pruning preserves nearly 99.8% of the original accuracy while reducing MAC operations and computation cycles. For SNN-oriented workloads, the proposed AND-type bitcell efficiently supports spike-weight multiplication with low switching activity, where the 2A2W configuration achieves accuracy close to the FP32 baseline across VGG-8, VGG-16, and ResNet-18 networks on CIFAR-10, CIFAR-100, and ImageNet-1K datasets. Compared to prior ADC-based ReRAM-CIM designs, the proposed architecture improves latency and energy efficiency by nearly 30-40% while maintaining robust operation under full PVT and ReRAM variability. Overall, E-ReCON provides a scalable, low-latency, and energy-efficient nvCIM platform for next-generation edge-AI, IoT, biomedical sensing, and neuromorphic applications.

Analog content-addressable memories (aCAMs) based on memristors provide a promising pathway toward energy-efficient large-scale associative computing for Edge AI and embedded intelligence applications. They have been successfully applied to decision-tree inference and extend the capabilities of compute-in-memory (CIM) architectures beyond conventional vector-matrix multiplication. However, conventional designs such as the 6T2M architecture suffer from static search power, limited voltage gain, and pronounced match-line crosstalk, constraining analog precision and scalability. We introduce a strong-arm latched memristor (SALM) aCAM cell that replaces static voltage division with a dynamic current-race comparator, enabling high regenerative gain, intrinsic result latching, and near-zero static search power. Compared to 6T2M, SALM reduces read energy by 33% at identical latency while eliminating the gain and crosstalk limitations that prevent 6T2M from scaling to large arrays. SALM further enables scalable sequential and parallel latch sharing, and a dataset-aware optimization framework exposes an explicit energy-latency tradeoff, achieving up to 50% energy reduction at 3x latency across representative workloads. To enable architectural exploration, we develop a circuit-accurate behavioral model derived from SPICE lookup tables in 22 nm FD-SOI technology, capturing match-line dynamics and crosstalk. Integrated into the X-TIME decision-tree compiler, this framework demonstrates that SALM maintains near-software accuracy for high-dimensional datasets, whereas baseline designs degrade due to limited gain and cumulative crosstalk.

Large language models (LLMs) with mixture-of-experts (MoE) architectures achieve remarkable scalability by sparsely activating a subset of experts per token, yet their frequent expert switching creates memory bandwidth bottlenecks that compute-in-memory (CIM) architectures are well-suited to mitigate. However, analog CIM systems suffer from inherent hardware imperfections that perturb stored weights, and its negative impact on MoE-based LLMs in noisy CIM environments remains unexplored. In this work, we present the first systematic investigation of MoE-based LLMs under noise model calibrated with real chip measurements, revealing that hardware noise critically disrupts expert load balance and renders clean-trained routing decisions consistently suboptimal. Based on these findings, we propose ROMER, a post-training calibration framework that (1) replaces underactivated experts with high-frequency ones to restore load balance, and (2) recalibrates router logits via percentile-based normalization to stabilize routing under noise. Extensive experiments across multiple benchmarks demonstrate that ROMER achieves up to 58.6%, 58.8%, and 59.8% reduction in perplexity under real-chip noise conditions for DeepSeek-MoE, Qwen-MoE, and OLMoE, respectively, establishing its effectiveness and generalizability across diverse MoE architectures.

Hardware-aware training (HAT) is widely used to improve the robustness of neural networks on non-ideal AI accelerators, such as analog in-memory computing (IMC) systems. However, not all hardware-induced distortions are equally compensable by training. This paper presents a diagnostic framework that models hardware non-idealities as structured perturbations of the forward operator and evaluates their compatibility with gradient-based optimization. We analyze six representative perturbation classes--read noise, variability, drift, stuck-at faults, IR-drop, and ADC discretization--and identify three key diagnostics: gradient expectation consistency, bounded gradient variance, and non-degenerate sensitivity. Our results show a clear separation between perturbations that can be compensated by HAT and those that consistently break optimization. This provides practical guidance for hardware-software co-design, clarifying which non-idealities can be addressed at the training level and which require circuit-, architecture-, or calibration-level mitigation. This study should be interpreted as a controlled empirical analysis under vanilla forward-perturbation HAT, rather than as a universal theory of hardware-aware training.

All current LLM serving systems place the GPU at the center, from production-level attention-FFN disaggregation to NVIDIA's Rubin GPU-LPU heterogeneous platform. Even academic PIM/PNM proposals still treat the GPU as the central hub for cross-device communication. Yet the GPU's compute-rich architecture is fundamentally mismatched with the memory-bound nature of decode-phase attention, inflating serving latency while wasting power and die area on idle compute units. The problem is compounded as reasoning and agentic workloads push context lengths toward one million tokens, making attention latency the primary user-facing bottleneck. To address these inefficiencies, we present AMMA, a multi-chiplet, memory-centric architecture for low-latency long-context attention. AMMA replaces GPU compute dies with HBM-PNM cubes, roughly doubling the available memory bandwidth to better serve memory-bound attention workloads. To translate this bandwidth into proportional performance gains, we introduce (i) a logic-die microarchitecture that fully exploits per-cube internal bandwidth for decode attention under a minimal power and area budget, (ii) a two-level hybrid parallelism scheme, and (iii) a reordered collective flow that reduces intra-chip die-to-die communication overhead. We further conduct a design-space exploration over per-cube compute power and intra-chip D2D link bandwidth, providing actionable guidance for hardware designers. Evaluations show that AMMA achieves 15.5X lower attention latency and 6.9X lower energy consumption compared with the NVIDIA H100.

In-memory computing (IMC) is a paradigm that enables neural network inference by computing analog matrix-vector multiplications (MVM) directly in memory crossbar arrays, with the potential for energy efficiency gains over conventional von Neumann architectures. In this work we present a simulation framework for N-ary crossbar architectures that retrieves MVM results with minimal implementation assumptions. The XOR and MNIST classification tasks were successfully inferred using a simulated crossbar array of (4x4) 4-states magnetic tunnel junctions (MTJ). MNIST accuracy reached 93.56% (vs. 97.56% software baseline). PCA dimensionality reduction was shown to drastically lower the number of required operations and improve the software baseline, for only a modest reduction in crossbar inference accuracy. We identified weight quantization as the primary error source, and studied its impact alongside systematic non-idealities and random noise. We find that cell-specific random noise is less detrimental than systematic errors due to averaging across the array. Finally, we demonstrate an optimal number of states per cell that balances quantization error against resistance state resolution to minimize total MVM error.

Speculative decoding enhances the inference efficiency of large language models (LLMs) by generating drafts using a small draft language model (DLM) and verifying them in batches with a large target language model (TLM). However, adaptive drafting inference on a mobile single-NPU-PIM system faces idle overhead in traditional operator-level synchronous execution and wasted computation in asynchronous execution due to fluctuations in draft length. This paper introduces AHASD, a task-level asynchronous mobile NPU-PIM heterogeneous architecture for speculative decoding. Notably, AHASD achieves parallel drafting on the PIM and verification on a single NPU through task-level DLM-TLM decoupling and specifically, it incorporates Entropy-History-Aware Drafting Control and Time-Aware Pre-Verification Control to dynamically manage adaptive drafting algorithm execution and pre-verification timing, suppressing invalid drafting based on low-confidence drafts. Additionally, AHASD integrates Attention Algorithm Units and Gated Task Scheduling Units within LPDDR5-PIM to enable attention link localization and sub-microsecond task switching on the PIM side. Experimental results for different LLMs and adaptive drafting algorithms show that AHASD achieves up to 4.2$\times$ in throughput and 5.6$\times$ in energy efficiency improvements over a GPU-only baseline, and 1.5$\times$ in throughput and 1.24$\times$ in energy efficiency gains over the state-of-the-art GPU+PIM baseline, with hardware overhead below 3% of the DRAM area.

Processing-in-Memory (PIM) architectures offer a promising solution to the memory bottlenecks in data-intensive machine learning, yet often overlook the growing challenge of activation memory footprint. Conventional PIM approaches struggle with massive KV cache sizes generated in long-context scenarios by Transformer-based models, frequently exceeding PIM's limited memory capacity, while techniques like sparse attention can conflict with PIM's need for data locality. Existing PIM approaches and quantization methods are often insufficient or poorly suited for leveraging the unique characteristics of activations. This work identifies an opportunity for PIM-specialized activation quantization to enhance bandwidth and compute efficiency. We explore clustering-based vector quantization approaches, which align well with activation characteristics and PIM's internal bandwidth capabilities. Building on this, we introduce AQPIM, a novel PIM-aware activation quantization framework based on Product Quantization (PQ), optimizing it for modern Large Language Models (LLMs). By performing quantization directly within memory, AQPIM leverages PIM's high internal bandwidth and enables direct computation on compressed data, significantly reducing both memory footprint and computational overhead for attention computation. AQPIM addresses PQ's accuracy challenges by introducing several algorithmic optimizations. Evaluations demonstrate that AQPIM achieves significant performance improvements, drastically reducing of GPU-CPU communication that can account for 90$\sim$98.5% of decoding latency, together with 3.4$\times$ speedup over a SOTA PIM approach.