arXiv subjects

Franck Cappello

Publications and source records attributed to Franck Cappello.

At least 19 recordsLinked to original sources

arXiv2026

KVStream: Adaptive Parallel Storage Management for KV Caching in Large-Scale LLM Serving

Long-context LLM serving for agentic, coding, and document workloads reuses prefixes whose key-value (KV) cache outgrows GPU and host memory, putting storage I/O on the critical path of the inference step for retrieving prefixes. A single node-local NVMe absorbs all retrievals, while the parallel file system or cloud file store that nodes already mount carries almost no inference traffic. We cast prefix retrieval as aggregate-bandwidth scheduling across these tiers and present KVStream, a KV cache engine that batches a step's reads, defers writes behind the forward pass, replicates each chunk on every tier, and lets each tier adaptively claim the next unread chunks once it is free. Our performance model shows this routing attains the aggregate-bandwidth bound without prediction or tuning, whereas any static split misses it by a margin that grows as per-tier bandwidths drift or fluctuate. It also predicts that the time-to-first-token (TTFT) gain grows with load, as faster retrieval drains the queue, and disappears when host memory or arrival order keeps reuse in cache. Plugged into LMCache and evaluated on four models, six datasets, and up to 128 GPUs, KVStream retrieves within 5% of the bound, delivers up to 1.49$\times$ higher offline throughput than LMCache, and lowers mean TTFT by up to 6.1$\times$ under bursty arrivals.

cs.DC↗

arXiv2026

PLEDGE: Progressive Lossy Error-Bounded Compression of 3D Gaussian Splatting

Existing 3D Gaussian Splatting (3DGS) codecs report reconstruction quality after compression, and give users little control over the distortion of individual primitives. We present PLEDGE, a training-free, progressive, error-bounded codec that translates four user-declared tolerances into per-primitive quantization bounds: screen-space position, footprint deformation due to rotation and scale, view-dependent color, and opacity. The bounds apply throughout a declared camera domain, including previously unseen viewpoints. Position and opacity are bounded directly; footprint deformation is bounded under the rasterizer's affine projection model, and color is bounded over all viewing directions. PLEDGE organizes quantized residuals into a progressive bitstream with explicit error bounds for every valid prefix. Experiments across thirteen scenes verify all declared per-mechanism bounds and demonstrate competitive rate-distortion performance. The four tolerances incur substantially different rate costs, so users can spend rate where their application needs it.

cs.CV↗

arXiv2026

LOCAA: An Agentic System for Automated Lossy Compressor Tuning

Large-scale scientific simulations generate substantial data volumes, making lossy compression essential for reducing storage and data movement costs. However, users configure compressors through numerical error bounds (EBs) while often evaluating results using quality metrics and end-to-end performance. Because the relationship between an EB and these outcomes varies across datasets and compressors, identifying a suitable configuration typically requires exhaustive search, which can be time-consuming and computationally demanding. We present LOCAA, an Large language model-based scientific lossy compression auto-tuning agent that performs compression-in-the-loop search using tool-integrated execution, compressor-aware guidance, and persistent memory. LOCAA supports user-defined objectives and constraints without requiring a specialized search strategy. We evaluate LOCAA across three scientific applications, five compressors, and three use cases: fixed-ratio compression, compression tuning under multiple quality constraints, and compression tuning under quality and time constraints. For fixed-ratio search across twelve fields from two applications, LOCAA requires 1.98x fewer evaluation trials than binary search and 5.03x fewer than FRaZ on average. For compression ratio maximization under joint Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure constraints across six fields from the Community Earth System Model application, LOCAA reduces the average evaluation trials from 61.5 using binary search to 17, corresponding to a 72.4% reduction. Persistent memory further reduces the average number of trials by 27.9% across multiple timesteps of the same field. These results demonstrate the potential of tool-augmented LLM agents to provide flexible and efficient compressor tuning across diverse scientific data and user-defined objectives.

cs.DC↗

arXiv2026

Measure Before You Manage: Evaluating Agent Working Memory in Coding Agents

Agent working memory is heterogeneous. Objects such as instructions, artifacts, tool outputs, and agent-generated state play different semantic roles and exhibit different size, retention, and representation profiles. Recent work has begun to explore memory-management mechanisms that account for such heterogeneity. This work focuses on semantic heterogeneity and studies how it should shape the management and evaluation of working memory in coding agents. Across 55 archived coding-agent trajectories, we find that semantically different working-memory objects exhibit distinct retention and compression behavior. This heterogeneity motivates semantically informed memory management. We study two semantically informed strategies: an object-aware compression policy and a retrieval-based policy. Their evaluation shows that calibration gains may not transfer to held-out tasks, and that equal token budgets do not imply equal delivered context or management cost. A real-system replay further exposes serving limits that nominal budgets alone do not capture. Together, these results show why semantic structure matters for agent working memory and why evaluating memory-management strategies requires more than a nominal token budget. We organize these lessons into four levels: stored state, delivered context, management work, and task or process outcome.

cs.AI↗

arXiv2026

Report of the 2026 Workshop on Next-Generation Ecosystems for Scientific Computing: Harnessing Community, Software, and AI for Cross-Disciplinary Team Science

Scientific computing is undergoing rapid transformation as advances in artificial intelligence, heterogeneous computing, automation, and data-intensive research reshape not only computational tools but also the institutions, workforce models, and collaborative practices that support scientific discovery. This report synthesizes insights from the 2026 Workshop on Next-Generation Ecosystems for Scientific Computing, the second in a three-year series focused on strengthening scientific computing ecosystems through socio-technical co-design. Workshop discussions identified four interdependent strategic themes: software ecosystems for AI-enabled scientific discovery; trust, validation, and traceability; human-AI teaming and paradigm shifts; and workforce, pedagogy, and governance. The report translates these themes into eight priorities for community action spanning shared research infrastructure, trust and traceability, user experience, human-AI teaming, workforce development, cross-sector coordination, stewardship and sustainability, and evaluation of scientific value. Together, these priorities outline directions for building scientific computing ecosystems that remain trustworthy, sustainable, innovative, and resilient as AI assumes a growing role in scientific work.

cs.CE↗

arXiv2026

FZ-VIS: A Visual Analytics Framework for Quantities-of-Interest-Aware Scientific Lossy Compression

Modern scientific simulations generate massive volumes of data, making lossy compression essential for efficient storage and transmission. However, preserving critical quantities of interest (QoIs) under lossy compression is inherently data- and task-dependent, requiring domain scientists to navigate complex trade-offs between compression ratio and data fidelity. Exploring these trade-offs often involves large design and evaluation spaces, motivating human-in-the-loop approaches that combine interactive exploration with quantitative analysis. To address this challenge, we present FZ-VIS, an interactive framework for human-in-the-loop feature-oriented lossy compression design and visual analytics. FZ-VIS provides a web-based interface for rapidly generating and comparing compression configurations, along with integrated visualization tools for assessing reconstruction fidelity and QoI preservation through both visual inspection and quantitative metrics. We demonstrate the utility of FZ-VIS through case studies involving three representative user groups: novice users selecting compression methods, compressor developers examining internal pipeline behavior, and domain scientists investigating feature preservation. The case studies show how FZ-VIS helps users efficiently navigate complex design spaces and make informed decisions that balance compression performance with application-specific QoI requirements.

cs.HC↗

arXiv2026

Preserving Clusters in Error-Bounded Lossy Compression of Scientific Particle Data

Scientific particle simulations in cosmology, molecular dynamics, and fluid dynamics produce large-scale datasets whose storage, movement, and analysis increasingly rely on lossy compression. However, existing compressors typically bound only pointwise position errors, providing no guarantee on the fidelity of structures derived from particle coordinates, such as single-linkage clustering (also known as Friends-of-Friends algorithm), where clusters are connected components of a proximity graph formed by linking particle pairs within a distance threshold. Even small coordinate perturbations near this threshold can break true links or create false links, thereby splitting or merging entire clusters. We propose a compressor-independent correction technique for preserving single-linkage cluster membership under lossy compression. Our method operates on reconstructed outputs from off-the-shelf compressors such as SZ3, ZFP, Draco, and LCP, and stores a compact corrective edit stream. Our key observation is that cluster-membership queries depend on connected components rather than the complete set of proximity links. Based on this observation, we introduce three constraint-selection modes, vulnerable-pair, safe-component, and halo-forest, that progressively reduce the constraints enforced during correction. Projected gradient descent then corrects the reconstructed coordinates to eliminate the selected violations while respecting the original pointwise error bound. Experiments on cosmology, molecular dynamics, and fluid dynamics datasets with single-GPU and distributed-memory implementations show that our method preserves cluster membership while improving compression ratio by up to 4$\times$ and maintains competitive end-to-end throughput compared to the same base compressors configured with sufficiently tight error bounds to preserve clustering.

cs.LG↗

arXiv2026

SPARe: Stacked Parallelism with Adaptive Reordering for Fault-Tolerant LLM Pretraining Systems with 100k+ GPUs

In large-scale LLM pre-training systems with 100k+ GPUs, failures become the norm rather than the exception, and restart costs can dominate wall-clock training time. However, existing fault-tolerance mechanisms are largely unprepared for this restart-dominant regime. To address this challenge, we propose SPARe - Stacked Parallelism with Adaptive Reordering - a fault-tolerance framework that masks node failures during gradient synchronization by stacking redundant data shards across parallelism groups and adaptively reordering execution. SPARe achieves availability comparable to traditional replication while maintaining near-constant computation overhead of only 2~3x, even under high redundancy where traditional replication would require linearly inflating overhead. We derive closed-form expressions for endurable failure count and computation overhead, validate them via SimGrid-based discrete-event simulation, and jointly optimize redundancy and checkpointing to minimize time-to-train. At extreme scale with up to 600k GPUs, SPARe reduces time-to-train by 40~50% compared to traditional replication.

cs.DC↗

arXiv2026

ReCoVer: Resilient LLM Pre-Training System via Fault-Tolerant Collective and Versatile Workload

Pre-training large language models on massive GPU clusters has made hardware faults routine rather than rare, driving the need for resilient training systems. Yet existing frameworks either focus on specific parallelism schemes or risk drifting away from a failure-free training trajectory. We propose ReCoVer, a resilient LLM pre-training system that upholds a single invariant: each iteration keeps the number of microbatches constant, ensuring per-iteration gradients remain stochastically equivalent to a failure-free run. The framework is organized as three decoupled protocol layers: (1) Fault-tolerant collectives that isolate faults from propagating across replicas; (2) in-step fine-grained recovery that preserves intra-iteration progress and prevents gradient corruption; (3) versatile-workload policy that dynamically redistributes microbatch quotas across the survivors. The design is parallelism-agnostic, integrating directly with both 3D parallelism and Hybrid Sharded Data Parallel (HSDP) as a drop-in substrate. We evaluate our implementation on end-to-end pre-training tasks for up to 512 GPUs, ReCoVer successfully preserves the training trajectory from a failure-free reference despite of 256 GPUs lost spread across the run. For comparison with checkpoint-and-restart baselines, ReCoVer demonstrates $2.23\times$ higher effective throughput after successive failures. This advantage results in ReCoVer processing 74.9% more tokens at 234 GPU-hours, with the gap widening as the training prolongs.

cs.DC↗

arXiv2026

BOOST: BOttleneck-Optimized Scalable Training Framework for Low-Rank Large Language Models

The scale of transformer model pre-training is constrained by the increasing computation and communication cost. Low-rank bottleneck architectures offer a promising solution to significantly reduce the training time and memory footprint with minimum impact on accuracy. Despite algorithmic efficiency, bottleneck architectures scale poorly under standard tensor parallelism. Simply applying 3D parallelism designed for full-rank methods leads to excessive communication and poor GPU utilization. To address this limitation, we propose BOOST, an efficient training framework tailored for large-scale low-rank bottleneck architectures. BOOST introduces a novel Bottleneck-aware Tensor Parallelism, and combines optimizations such as online-RMSNorm, linear layer grouping, and low-rank activation checkpointing to achieve end-to-end training speedup. Evaluations on different low-rank bottleneck architectures demonstrate that BOOST achieves 1.46-1.91$\times$ speedup over full-rank model baselines and 1.87-2.27$\times$ speedup over low-rank model with naively integrated 3D parallelism, with improved GPU utilization and reduced communication overhead.

cs.LG↗

arXiv2026

Deep Optimizer States: Towards Scalable Training of Transformer Models Using Interleaved Offloading

Transformers and large language models~(LLMs) have seen rapid adoption in all domains. Their sizes have exploded to hundreds of billions of parameters and keep increasing. Under these circumstances, the training of transformers is very expensive and often hits a ``memory wall'', i.e., even when using 3D parallelism (pipeline, tensor, data) and aggregating the memory of many GPUs, it is still not enough to hold the necessary data structures (model parameters, optimizer state, gradients, activations) in GPU memory. To compensate, state-of-the-art approaches offload the optimizer state, at least partially, to the host memory and perform hybrid CPU-GPU computations. However, the management of the combined host-GPU memory is often suboptimal and results in poor overlapping between data movements and computations. This leads to missed opportunities to simultaneously leverage the interconnect bandwidth and computational capabilities of CPUs and GPUs. In this paper, we leverage a key observation that the interleaving of the forward, backward, and update phases generates fluctuations in the GPU memory utilization, which can be exploited to dynamically move a part of the optimizer state between the host and the GPU memory at each iteration. To this end, we design and implement Deep Optimizer States, a novel technique to split the LLM into subgroups, whose update phase is scheduled on either the CPU or the GPU based on our proposed performance model that addresses the trade-off between data movement cost, acceleration on the GPUs vs the CPUs, and competition for shared resources. We integrate our approach with DeepSpeed and demonstrate 2.5$\times$ faster iterations over state-of-the-art approaches using extensive experiments.

cs.LG↗

arXiv2026

Enabling Homomorphic Analytical Operations on Compressed Scientific Data with Multi-stage Decompression

Error-controlled lossy compressors have been widely used in scientific applications to reduce the unprecedented size of scientific data while keeping data distortion within a user-specified threshold. While they significantly mitigate the pressure for data storage and transmission, they prolong the time to access the data because decompression is required to transform the binary compressed data into meaningful floating-point numbers. This incurs noticeable overhead for common analytical operations on scientific data that extract or derive useful information, because the time cost of the operations could be much lower than that of decompression. In this work, we design an error-controlled lossy compression and analytical framework that features multi-stage decompression and homomorphic analytical operation algorithms on intermediate decompressed data for reduced data access latency. Our contributions are threefold. (1) We abstract a generic compression pipeline with partial decompression to multiple intermediate data representations and implement four instances based on state-of-the-art high-throughput scientific data compressors. (2) We carefully design homomorphic algorithms to enable direct operations on intermediate decompressed data for three types of analytical operations on scientific data. (3) We evaluate our approach using five real-world scientific datasets. Experimental evaluations demonstrate that our method achieves significant speedups when performing analytical operations on compressed scientific data across all three targeted analytical operation types.

cs.DB↗

arXiv2026

Rate-Distortion Bounds for Heterogeneous Random Fields on Finite Lattices

Since Shannon's foundational work, rate-distortion theory has defined the fundamental limits of lossy compression. Classical results, derived for memoryless and stationary ergodic sources in the asymptotic regime, have shaped both transform and predictive coding architectures, as well as practical standards such as JPEG. Finite-blocklength refinements, initiated by the non-asymptotic achievability and converse bounds of Kostina and Verdu, provide precise characterizations under excess-distortion probability constraints, but primarily for memoryless or statistically homogeneous models. In contrast, error-bounded practical lossy compressors for scientific computing, such as SZ, ZFP, MGARD, and SPERR, are designed for finite, high-dimensional, spatially correlated, and statistically heterogeneous random fields. These compressors partition data into fixed-size tiles that are processed independently, making tile size a central architectural constraint. Structural heterogeneity, finite lattice effects, and tiling constraints are not addressed by existing finite-blocklength analyses. This paper introduces a finite-blocklength rate-distortion framework for heterogeneous random fields on finite lattices, explicitly accounting for the tile-based architectures used in high-performance scientific compressors. The field is modeled as piecewise homogeneous with regionwise stationary second-order statistics, and tiling constraints are incorporated directly into the source model. Under an excess-distortion probability criterion, we establish non-asymptotic achievability, converse bounds and derive a second-order expansion that quantifies the impact of spatial correlation, region geometry, heterogeneity, and tile size on the rate and dispersion.

cs.IT↗

arXiv2026

Mitigating Artifacts in Pre-quantization Based Scientific Data Compressors with Quantization-aware Interpolation

Error-bounded lossy compression has been regarded as a promising way to address the ever-increasing amount of scientific data in today's high-performance computing systems. Pre-quantization, a critical technique to remove sequential dependency and enable high parallelism, is widely used to design and develop high-throughput error-controlled data compressors. Despite the extremely high throughput of pre-quantization based compressors, they generally suffer from low data quality with medium or large user-specified error bounds. In this paper, we investigate the artifacts generated by pre-quantization based compressors and propose a novel algorithm to mitigate them. Our contributions are fourfold: (1) We carefully characterize the artifacts in pre-quantization based compressors to understand the correlation between the quantization index and compression error; (2) We propose a novel quantization-aware interpolation algorithm to improve the decompressed data; (3) We parallelize our algorithm in both shared-memory and distributed-memory environments to obtain high performance; (4) We evaluate our algorithm and validate it with two leading pre-quantization based compressors using five real-world datasets. Experiments demonstrate that our artifact mitigation algorithm can effectively improve the quality of decompressed data produced by pre-quantization based compressors while maintaining their high compression throughput.

cs.DC↗

arXiv2026

TopoSZp: Lightweight Topology-Aware Error-controlled Compression for Scientific Data

Error-bounded lossy compression is essential for managing the massive data volumes produced by large-scale HPC simulations. While state-of-the-art compressors such as SZ and ZFP provide strong numerical error guarantees, they often fail to preserve topological structures (example, minima, maxima, and saddle points) that are critical for scientific analysis. Existing topology-aware compressors address this limitation but incur substantial computational overhead. We present TopoSZp, a lightweight, topology-aware, error-controlled lossy compressor that preserves critical points and their relationships while maintaining high compression and decompression performance. Built on the high-throughput SZp compressor, TopoSZp integrates efficient critical point detection, local ordering preservation, and targeted saddle point refinement, all within a relaxed but strictly enforced error bound. Experimental results on real-world scientific datasets show that TopoSZp achieves 3 to 100 times fewer non-preserved critical points, introduces no false positives or incorrect critical point types, and delivers 100 to 10000 times faster compression and 10 to 500 times faster decompression compared to existing topology-aware compressors, while maintaining competitive compression ratios.

cs.DC↗

arXiv2026

DataStates-LLM: Scalable Checkpointing for Transformer Models Using Composable State Providers

The rapid growth of Large Transformer-based models, specifically Large Language Models (LLMs), now scaling to trillions of parameters, has necessitated training across thousands of GPUs using complex hybrid parallelism strategies (e.g., data, tensor, and pipeline parallelism). Checkpointing this massive, distributed state is critical for a wide range of use cases, such as resilience, suspend-resume, investigating undesirable training trajectories, and explaining model evolution. However, existing checkpointing solutions typically treat model state as opaque binary blobs, ignoring the ``3D heterogeneity'' of the underlying data structures--varying by memory location (GPU vs. Host), number of ``logical'' objects sharded and split across multiple files, data types (tensors vs. Python objects), and their serialization requirements. This results in significant runtime overheads due to blocking device-to-host transfers, data-oblivious serialization, and storage I/O contention. In this paper, we introduce DataStates-LLM, a novel checkpointing architecture that leverages State Providers to decouple state abstraction from data movement. DataStates-LLM exploits the immutability of model parameters during the forward and backward passes to perform ``lazy'', non-blocking asynchronous snapshots. By introducing State Providers, we efficiently coalesce fragmented, heterogeneous shards and overlap the serialization of metadata with bulk tensor I/O. We evaluate DataStates-LLM on models up to 70B parameters on 256 A100-40GB GPUs. Our results demonstrate that DataStates-LLM achieves up to 4$\times$ higher checkpointing throughput and reduces end-to-end training time by up to 2.2$\times$ compared to state-of-the-art solutions, effectively mitigating the serialization and heterogeneity bottlenecks in extreme-scale LLM training.

cs.DC↗

arXiv2026

pMSz: A Distributed Parallel Algorithm for Correcting Extrema and Morse Smale Segmentations in Lossy Compression

Lossy compression, widely used by scientists to reduce data from simulations, experiments, and observations, can distort features of interest even under bounded error. Such distortions may compromise downstream analyses and lead to incorrect scientific conclusions in applications such as combustion and cosmology. This paper presents a distributed and parallel algorithm for correcting topological features, specifically, piecewise linear Morse Smale segmentations (PLMSS), which decompose the domain into monotone regions labeled by their corresponding local minima and maxima. While a single GPU algorithm (MSz) exists for PLMSS correction after compression, no methodology has been developed that scales beyond a single GPU for extreme scale data. We identify the key bottleneck in scaling PLMSS correction as the parallel computation of integral paths, a communication-intensive computation that is notoriously difficult to scale. Instead of explicitly computing and correcting integral paths, our algorithm simplifies MSz by preserving steepest ascending and descending directions across all locations, thereby minimizing interprocess communication while introducing negligible additional storage overhead. With this simplified algorithm and relaxed synchronization, our method achieves over 90% parallel efficiency on 128 GPUs on the Perlmutter supercomputer for real world datasets.

cs.DC↗

arXiv2026

FFCz: Fast Fourier Correction for Spectrum-Preserving Lossy Compression of Scientific Data

This paper introduces a novel technique to preserve spectral features in lossy compression based on a novel fast Fourier correction algorithm\added{ for regular-grid data}. Preserving both spatial and frequency representations of data is crucial for applications such as cosmology, turbulent combustion, and X-ray diffraction, where spatial and frequency views provide complementary scientific insights. In particular, many analysis tasks rely on frequency-domain representations to capture key features, including the power spectrum of cosmology simulations, the turbulent energy spectrum in combustion, and diffraction patterns in reciprocal space for ptychography. However, existing compression methods guarantee accuracy only in the spatial domain while disregarding the frequency domain. To address this limitation, we propose an algorithm that corrects the errors produced by off-the-shelf ``base'' compressors such as SZ3, ZFP, and SPERR, thereby preserving both spatial and frequency representations by bounding errors in both domains. By expressing frequency-domain errors as linear combinations of spatial-domain errors, we derive a region that jointly bounds errors in both domains. Given as input the spatial errors from a base compressor and user-defined error bounds in the spatial and frequency domains, we iteratively project the spatial error vector onto the regions defined by the spatial and frequency constraints until it lies within their intersection. We further accelerate the algorithm using GPU parallelism to achieve practical performance. We validate our approach with datasets from cosmology simulations, X-ray diffraction, combustion simulation, and electroencephalography demonstrating its effectiveness in preserving critical scientific information in both spatial and frequency domains.

cs.DC↗