Chips and semiconductors: research

Journal articles, academic blogs, think-tank reports and papers in this section, newest first. Back to the news.

arXiv (research papers) Guilherme Guedes

Hypnos: A Dual-Phase End-to-End Platform for Scalable Continual Learning at the Edge

Continual learning (CL) enables neural networks to adapt to non-stationary data streams by mitigating catastrophic forgetting, offering a path to robust, private, and adaptable edge devices. Yet, porting CL to real-world edge devices remains challenging for two reasons: (i) online operation, where sensor data arrives one sample at a time and the device must adapt on-the-fly, and (ii) the low footprint of edge devices, which makes traditional error backpropagation training…

Read at arXiv (research papers)

arXiv (research papers) Trevor McCourt

Reliable arithmetic circuits using less reliable (sub-digital) logic

The frequent static failure of CMOS logic gates at ultra-low voltages has historically been perceived as a strict barrier to extreme voltage scaling: this paper introduces a statistical circuit law proving this intuition wrong. We discovered a universal power law connecting gate and circuit static failure probabilities, revealing that circuit failure requires the simultaneous breakdown of an entire gate cluster rather than a single component. Because the required cluster…

Read at arXiv (research papers)

arXiv (research papers) Philipp Schmitz

TechReport: Implementing CHERI in gem5

Despite countless mitigation efforts, memory safety issues still prevail. The recently discovered security issues in PHP underline this fact. One of the most promising approaches to solving this problem is the concept of Capability Hardware Enhanced RISC Instructions (CHERI). There is a well-developed ecosystem of ISA simulators, compilers, operating systems, and hardware designs that support CHERI. However, the lack of a simulator modeling microarchitectural detail hinders…

Read at arXiv (research papers)

arXiv (research papers) Ziwen Wang

A Runtime System for Mitigating Crossbar Port Contention on SM-Partitioned GPUs

To avoid scheduling delays when running concurrent workloads on a GPU, Streaming Multiprocessor (SM) partitioning mechanisms such as CUDA Green Contexts restrict each workload to separate SMs. Despite this partitioning, concurrent workloads still significantly slow down latency-critical workloads, which recent studies attribute to shared L2 cache and memory contention. However, contention also occurs at the crossbar ports connecting each Graphics Processing Cluster (GPC) of…

Read at arXiv (research papers)

arXiv (research papers) Ziwen Wang

Dissecting an Unexplored SASS Control-Code Field: Barrier Exemption and Programmable Warp Priority

Every instruction in NVIDIA's native GPU ISA (SASS) carries a control code that guides hardware scheduling, and prior work has decoded its known fields to optimize GPU kernels. However, one control-code field remains unexplored: it appears over 500000 times in NVIDIA's closed-source libraries such as cuBLAS and cuDNN, yet nvcc never emits it when compiling user programs. This paper presents the first systematic study of this control-code field across GPU architectures from…

Read at arXiv (research papers)

arXiv (research papers) Yuchen Li

How to Turn an Attention Neural Network into an Interpretable Hard-Logic Predicate Circuit: Truth-Table Predicate Extraction and State-Machine Merging under Ship-of-Theseus Replacement

We convert a trained attention network as a whole, attention included, into a hard-logic predicate circuit that runs in its place. Every unit of the circuit is a 0/1 predicate on tokens, and every map between predicates is a minimised truth table. Short tables, such as the product of a query code and a key code, are minimised exactly, and wide feature-to-feature transitions become implication rules by noise-tolerant minimisation. Blocks are replaced one at a time from the…

Read at arXiv (research papers)

arXiv (research papers) Robin Geens

Hardware Acceleration of Vision-Language-Action Models with Weight Precision Rising over the Flow Steps

Vision-language-action (VLA) models such as GR00T control robots through an action head: a diffusion transformer that turns noise into a chunk of actions in a few flow-matching steps. Every step re-reads all of the head's weights from DRAM but applies each weight to only about fifty tokens. Inference is therefore memory-bound, and its latency and energy are largely determined by the total number of transferred weight bits. We show that later flow steps attenuate the…

Read at arXiv (research papers)

arXiv (research papers) Niccolò Lentini Security

PULSAR-V: I/O-Aware Post-Update Attestation for RISC-V

Post-update attestation proves that a device installed the intended firmware, not that the image stayed protected until the report was signed. In a multi-master SoC, a master outside the attested processor can corrupt update-critical memory. In contrast, the firmware identity, lifecycle state, and the CPU's protection evidence remain byte-identical to a healthy device's. We present PULSAR-V, a post-update attestation architecture which binds the authenticated update commit…

Read at arXiv (research papers)

arXiv (research papers) Keming Fan

ReCIM-HD: A 40 nm On-Device Learning Accelerator with Hyperdimensional Computing via Importance-Based Hybrid Digital-ReRAM CIM

ReCIM-HD is an on-device learning (ODL) accelerator that integrates gradient-free hyperdimensional computing (HDC) with resistive RAM (ReRAM)-based analog compute-in-memory (CIM). ReCIM-HD introduces three key techniques: (1) ultra-low-power HDC-based edge ODL without costly back-propagation operations; (2) non-volatile ReRAM-based storage for class hypervectors and CIM similarity search, enabling energy-efficient inference and power-gating during idle periods without data…

Read at arXiv (research papers)

arXiv (research papers) Vikash Kumar

Inference-Time Scaling in Agentic Hardware Verification Workflows: A Multi-Strategy Characterization Across Protocol Complexity

Inference-time scaling, allocating more compute during generation rather than pre-training, has demonstrated dramatic quality improvements for LLM reasoning tasks. Its behaviour in structured hardware verification code generation, where outputs must satisfy compiler correctness, protocol semantics, and functional coverage constraints simultaneously, is less understood. We present a systematic, simulator-grounded study of inference-time compute scaling for agentic UVM…

Read at arXiv (research papers)

arXiv (research papers) Tong Wu

Accelerating Distributed Digital Ising Machines

Ising machines, dedicated accelerators built around the dynamics of Ising-like spin models, have emerged as a leading platform for combinatorial optimization and probabilistic sampling. The problems that motivate them require millions of spins, far beyond the on-chip capacity of any single device, so a practical Ising machine must distribute its computation across many devices. Prior work reports throughput at a single operating point, leaving the binding resource…

Read at arXiv (research papers)

arXiv (research papers) Zhiheng Chen

HETEROVLM: Edge Multi-Request Vision-Language Model Serving on a Heterogeneous Versal SoC with AI Engines and Programmable Logic

Edge devices in vehicles and robots serve concurrent requests with different deadlines to a vision-language model (VLM), each mixing a compute-bound encoder and prefill with a memory-bound decode. GPU serving targets datacenter throughput with per-iteration host scheduling: an urgent request waits for the running iteration, and small-batch decode costs much energy per token. Prior accelerators on Versal AI Engine (AIE) SoCs map one model onto a fixed AIE dataflow graph for…

Read at arXiv (research papers)

arXiv (research papers) Peizhe Liu

Lodestone: Automated Semantic Discovery of Datacenter Hardware Specialization

Hardware specialization can improve datacenter efficiency. Unfortunately, identifying computations that justify dedicated hardware is difficult. Existing approaches analyze hot functions, instruction patterns, or overheads such as datacenter tax. They can miss fine-grained computations whose aggregate cost is spread across applications and libraries, especially when different pieces of code obscure the same underlying computation. We observe that seemingly unrelated code…

Read at arXiv (research papers)

arXiv (research papers) Shoma Yukawa

EEG2MCU: Sensing-to-Execution Co-Design for Wearable EEG Inference

Wearable EEG inference is constrained not only by model size, but also by the channel-dependent sensing front end that must process high-dimensional inputs before compact downstream computation. We present EEG2MCU, a sensing-to-execution co-design framework that treats electrode count as a hardware design variable through task-aware montage selection. EEG2MCU retains a compact time-frequency representation and generates only the INT8 regions required by each first-layer…

Read at arXiv (research papers)

arXiv (research papers) Sehyeon Lee

Towards Fair Resource Sharing in Agentic LLM Serving

As LLM agents increasingly execute complex, multi-step tasks on shared inference infrastructure, resource contention across the end-to-end LLM serving stack causes severe and uneven slowdowns across agents. Existing agentic serving systems primarily optimize throughput and latency, with comparatively limited attention to fairness among concurrent agents. Although fairness has been studied in conventional LLM serving at the token and request levels, these approaches do not…

Read at arXiv (research papers)

arXiv (research papers) Malak Gamal El-Din

BeFit: Shifting Complexity from Weights to Bernstein Activations for Resource-Constrained Sensing

Time-series sensing increasingly relies on neural networks, yet deployment on resource-constrained microcontrollers remains challenging. Aggressive model compression can reduce accuracy, while restricting weights to multiplication-free representations introduces further losses. We present BeFit, an end-to-end framework that shifts model capacity from weights to learnable Bernstein polynomial activations. BeFit (i) finds compact Bernstein-activated convolutional neural…

Read at arXiv (research papers)

arXiv (research papers) Qirui Zhang

Compute Before You Digitize: ADC-less Mixed-Signal Processing for Scalable Cryogenic Qubit Readout

We introduce DIRAC, an ADC-less in-cryostat qubit readout architecture that uses analog computing to redraw the analog-digital boundary. Directly operating on the front-end's analog output, passive complex mixers perform the equivalent of digital demodulation, ping-pong active-RC integrators average their outputs into down-sampled per-qubit traces, and a time-domain stage digitizes them, eliminating Nyquist-rate ADCs. A TX-RX co-design derives the mixers' local oscillators…

Read at arXiv (research papers)

Atlantic Council Jessie Yin Geopolitics

As the Nvidia scandal threatens to reignite trade tensions at Trump-Xi talks, Beijing still holds the rare earths

Washington’s push to close loopholes could invite renewed Chinese restrictions on rare earths, exposing a supply-chain vulnerability that the trade truce never resolved.

Read at Atlantic Council

arXiv (research papers) Yuanlin Wen

HardwareDx: A Foundational Graph Model System for Datacenter Machine Repairs

The rapid scale-up of AI infrastructure has made datacenter reliability a cornerstone of the cloud business economy. Since hardware failures in hyper-scale AI clusters can stall multi-million dollar training workloads, the downtime of individual machines can cause outsized economic losses. Minimizing machine downtime, therefore, is more important than ever. To address this, we present Hardware Diagnoser (HardwareDx), an AI-driven hardware diagnosis system that applies a…

Read at arXiv (research papers)

arXiv (research papers) Owen Sullivan

CADENCE: Learning Acyclic Gaits for Model-Based Quadruped Locomotion

Model-based quadruped controllers usually step on a cyclic gait, which lifts the legs in a fixed order on a fixed period, wherever the safe footholds are. We present CADENCE, a learned step planner with a small action space. Every 40 ms, it picks one of at most 65 discrete choices: no step, or one leg's step to a safe foothold. It also predicts a swing time. We train CADENCE by reinforcement learning on top of a convex model predictive controller (MPC). Nothing prescribes a…

Read at arXiv (research papers)

arXiv (research papers) Yizhuo Wu

StreamingDPD: A 1.28-GSa/s Instantaneous-Power-Gated Depthwise TCN Accelerator for Real-Time 320-MHz Digital Predistortion

State feedback between successive samples limits the streaming throughput of recurrent neural networks for digital predistortion (DPD). We present StreamingDPD, an accelerator that processes 320-MHz signals at 1.28 GSa/s. Its 207-parameter instantaneous-power-gated depthwise temporal convolutional network (IPG-DWTCN) uses $|x[n]|^2$ to gate feature channels and causal dilated depthwise convolutions to model power amplifier memory. The network maps to a spatially unrolled…

Read at arXiv (research papers)

arXiv (research papers) Wenbo Han Security

Dissecting the GPU Load/Store Unit and Exposing Its Security Risks

GPUs are the main platform for large-scale machine learning and power the largest supercomputers. Their load/store units (LSUs) often determine kernel performance and have been exploited in prior attacks. Yet what is known about the LSU is only a few isolated mechanisms and parameters, and even a single one has been enough for an attack. In particular, it is still unknown what a load costs, where loads contend, and even whether a warp can starve due to contention. Until each…

Read at arXiv (research papers)

arXiv (research papers) Patrick Iff

LASH-PAT: Deadlock-Free Adaptive Shortest-Path Routing for Inter-Chiplet Interconnects

With the slowdown of transistor scaling, chipletbased systems, and in particular chiplet-based wafer-scale integration, have emerged as a promising approach to sustain growth in computing capability per cost for modern processors and accelerators. In recent years, many specialized inter-chiplet interconnect topologies have been proposed. In these topologies, the simple, well-established, and deadlock-free dimension-order routing scheme is not applicable due to its reliance…

Read at arXiv (research papers)

arXiv (research papers) Ihor Lukianov

blaris: Windowless Random-Access LZSS for Constrained Microcontrollers

Embedded LZ77 decompressors (such as heatshrink and tamp) require an in-RAM sliding window that couples memory usage to compression ratio and incurs high latency on random-access reads. We present blaris, a lightweight, windowless LZSS decompressor for execute-in-place (XIP) random access on constrained microcontrollers. By partitioning tokens into bidirectionally traversable streams with power-of-two checkpoints, blaris resolves back-references directly in the compressed…

Read at arXiv (research papers)

arXiv (research papers) Weronika Hryniewska-Guzik

How well does graph superoptimization transfer? Adapting tensor-graph superoptimizer to a Triton-based NPU

For automatic optimization systems, performance portability across increasingly diverse AI accelerators remains a challenge. We investigate how well graph-level superoptimization transfers across architectures by adapting Mirage from its CUDA-oriented design to the Huawei Ascend 910B3 through a Triton-based compilation stack. We evaluate the adapted system on transformer-relevant workloads, including matrix multiplication, softmax, RMSNorm+MatMul, gated MLP, FNN, and…

Read at arXiv (research papers)

arXiv (research papers) Patrick Iff

PAISF: A Partially Adaptive In-Order Selection Function for Routing in On-Chip, Inter-Chiplet, and Wafer-Scale Networks

The growing number of compute elements per chip, along with advances such as chiplet systems and wafer-scale integration, drives the need for scalable and efficient routing algorithms for irregular topologies. A routing algorithm consists of a routing function (RF) that ensures deadlock freedom and a selection function (SF) that avoids congestion and, in some cases, guarantees in-order packet delivery. We propose two novel SFs for irregular topologies, ISF and PAISF, which…

Read at arXiv (research papers)

arXiv (research papers) Yuseon Choi

LaFlash: Capacity-Driven Acceleration for Coding Agent Workflows via QLC-HBF

As large language models (LLMs) evolve toward agentic AI, coding agents are emerging as memory-intensive workloads that repeatedly reuse persistent states, including contexts and KV caches. Emerging agentic workflows, including Loop/Graph Engineering, cause state amplification through frequent active/dormant transitions, increasing memory pressure. Evicted states incur costly recomputation upon reactivation, particularly under long-context, short-decode workloads. This…

Read at arXiv (research papers)

arXiv (research papers) Fernanda Zapata Bascuñán

Gateway-Free Kafka Streaming from FPGA-Based Detectors: A Bare-Metal Pipelined Producer on Zynq-7000

We present a bare-metal Apache Kafka producer implemented on a Zynq-7000 System-on-Chip (SoC) for event-driven data acquisition. Events are generated in Programmable Logic (PL), transferred to the Processing System (PS) through AXI4-Stream and AXI DMA, and published to a standard Kafka broker without an intermediate gateway computer. The PS constructs ProduceRequest messages and transmits them through the lwIP raw TCP/IP API. A configurable pipeline limits the number of…

Read at arXiv (research papers)

arXiv (research papers) Wenbo Zhu

Manufacturing Differential Sparsity: Planner-Hardware Co-Design for Efficient World Model Planning

JEPA-style latent world models let embodied agents plan toward new goals with model predictive control, but they are still too slow for real-life embodied control: sampling-based MPC repeats predictor inference tens of thousands of times per plan, and even on high-end GPUs one planning step can take tens of seconds. We observe substantial latent locality in this repeated-inference workload: the predictor maps close actions to close latents, and inside a quantized predictor…

Read at arXiv (research papers)

arXiv (research papers) Siqing Fu

KD-DSE: A Knowledge-Driven LLM Framework for Fast Microarchitecture Design Space Exploration

Domain-specific architectures have made design space exploration (DSE) a bottleneck of processor design, yet each candidate costs a full simulation, so exhaustive search is out of reach. Surrogate models need thousands of simulations before they predict reliably, and LLM agents that search directly hallucinate, lack cold-start architectural knowledge, and follow a fixed per-round policy. We present KD-DSE, a knowledge- driven LLM framework for fast microarchitecture DSE. KD…

Read at arXiv (research papers)

Older stories