Related papers: The Fast Linear Accelerator Modeling Engine for FR…
The demand for efficient large language model (LLM) inference has propelled the development of dedicated accelerators. As accelerators are vulnerable to hardware faults due to aging, variation, etc, existing accelerator designs often…
LiDAR super-resolution addresses the challenge of achieving high-quality 3D perception from cost-effective, low-resolution sensors. While recent transformer-based approaches like TULIP show promise, they remain limited to spatial-domain…
Multi-Model Federated Learning (MMFL) is an emerging direction in Federated Learning (FL) where multiple models are trained in parallel, generally on various datasets. Optimizing the models' accuracies and training times in the MMFL setting…
The recently developed Flexible Local Approximation MEthod (FLAME) produces accurate difference schemes by replacing the usual Taylor expansion with Trefftz functions -- local solutions of the underlying differential equation. This paper…
In the ever-evolving landscape of artificial intelligence (AI) and large language models (LLMs), handling and leveraging data effectively has become a critical challenge. Most state-of-the-art machine learning algorithms are data-centric.…
The Fermi National Accelerator Laboratory (Fermilab) Linac accepts 750 keV H- ions from the front end and accelerates them to 400 MeV for injection into the Booster rapid cycling synchrotron. Day-to-day drifts in the beam longitudinal…
In this paper, we present a novel closed-form model (CFM) for accurate and fast evaluation of nonlinear interference in modern ultrawideband coherent optical fiber communication systems. Starting from the Gaussian noise model (GN model),…
The Elastic Analysis Facility (EAF) hosted at Fermi National Accelerator Laboratory (Fermilab) is a platform being developed with the goal of providing a fast and efficient facility for physics analysis. As high-energy physics moves towards…
FPGA is appropriate for fix-point neural networks computing due to high power efficiency and configurability. However, its design must be intensively refined to achieve high performance using limited hardware resources. We present an…
We present a fast and accurate analytical method for fluorescence lifetime imaging microscopy (FLIM) using the extreme learning machine (ELM). We used extensive metrics to evaluate ELM and existing algorithms. First, we compared these…
We summarize the current state of a concept for muon acceleration aimed at a future Neutrino Factory. The main thrust of these studies was to reduce the overall cost while maintaining performance by exploring the interplay between the…
In this paper, we propose FusionCIM, an operator-fusion-driven compute-in-memory (CIM) accelerator architecture for efficient and scalable LLM inference, with three key innovations: (1) a hybrid CIM pipeline architecture that maps QKT…
Beam matching is a common technique that is routinely employed in accelerator design with the aim of minimizing beam losses and preservation of beam brightness. Despite being widely used, a full theoretical understanding of beam matching in…
For decades, supercritical flame simulations incorporating detailed chemistry and real-fluid transport have been limited to millions of cells, constraining the resolved spatial and temporal scales of the physical system. We optimize the…
This paper proposes FAST-Calib, a fast and user-friendly LiDAR-camera extrinsic calibration tool based on a custom-made 3D target. FAST-Calib supports both mechanical and solid-state LiDARs by leveraging an efficient and reliable edge…
The FLARE code is a magnetic mesh generator that is integrated within a suite of tools for the analysis of the magnetic geometry in toroidal fusion devices. A magnetic mesh is constructed from field line segments and permits fast…
Among the algorithms that are likely to play a major role in future exascale computing, the fast multipole method (FMM) appears as a rising star. Our previous recent work showed scaling of an FMM on GPU clusters, with problem sizes in the…
Continuous-time Consistency Models (CMs) promise efficient few-step generation but face significant challenges with training instability. We argue this instability stems from a fundamental conflict: Training the network exclusively on a…
In this paper, a self-consistent transverse beam dynamics framework is demonstrated, that incorporates acceleration into the transverse beam dynamics studies for a proton linac machine. Two focusing schemes are developed and discussed; the…
Processing in-memory (PIM) is promising to accelerate neural networks (NNs) because it minimizes data movement and provides large computational parallelism. Similar to machine learning accelerators, application mapping, which determines the…