https://techgazette.com.br/research/#systems This week's papers in Systems and databases, each with what its authors did and found, in sentences from its own abstract. en 180 Sat, 17 Oct 2026 12:00:00 GMT https://dl.acm.org/doi/10.1145/3859287 https://dl.acm.org/doi/10.1145/3859287 Sat, 17 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Did: Serverless workflows are commonly modeled as directed acyclic graphs (DAGs) composed of fine-grained functions. https://arxiv.org/abs/2610.22310 https://arxiv.org/abs/2610.22310 Thu, 15 Oct 2026 17:50:00 GMT arXiv (systems & security) cs.DB Systems and databases arXiv (systems & security) cs.DB — Did: To this end, we propose TraHist, a histogram-based estimator approximating density-aligned partitioning via efficient grid-based propagation. Found: Experiments on large-scale real-world datasets show that TraHist reduces estimation error by over 20x relative to the strongest competing estimator in query regimes where trajectory fragmentation is a dominant source of estimation error, while maintaining sub-millisecond query latency and competitive construction cost. By Yujeong Jo, Incheol Baek, Changbeom Shim, Yon Dohn Chung. https://arxiv.org/abs/2610.22230 https://arxiv.org/abs/2610.22230 Thu, 15 Oct 2026 17:09:00 GMT arXiv cs.DC Systems and databases arXiv cs.DC — Did: This paper presents our methods for optimizing alternative accelerators, using AMD Instinct MI350X versus NVIDIA B200 for recommendation model training as a case study. Found: Compared to the directly ported baseline, these methods deliver up to 1.5x improvement in end-to-end training throughput. By Haoyu Zhang, Arshad Rahman, Adnan Aziz, Ajith Prabhudev, Amin Farmahini, Ben Rhoads. https://arxiv.org/abs/2610.22172 https://arxiv.org/abs/2610.22172 Thu, 15 Oct 2026 16:31:00 GMT arXiv (systems & security) cs.DC Systems and databases arXiv (systems & security) cs.DC — Did: In this work, we propose EdgeFlow, a novel hybrid KV cache migration scheme that integrates direct KV cache transmission with activation-based recomputation, leveraging parallel computation and communication to minimize migration time. Found: Extensive experiments on a real-world distributed LLM inference platform demonstrate that EdgeFlow reduces migration time by up to 45%, significantly enhancing the robustness and efficiency of distributed LLM inference in collaborative edge computing. By Mingjin Zhang, Jiannong Cao, Tao Wu, Xiangchun Chen, Ne Wang. https://arxiv.org/abs/2610.22035 https://arxiv.org/abs/2610.22035 Thu, 15 Oct 2026 15:10:00 GMT arXiv (systems & security) cs.DC Systems and databases arXiv (systems & security) cs.DC — Did: In this paper, we propose Heterogeneous Pipeline Parallelism (HPP), a novel approach enabling efficient and fine-grained data balance across heterogeneous pipelines. Found: Evaluations demonstrate ETH achieves up to 2x training speedup while reducing configuration generation overhead by up to 56x compared to state-of-the-art systems. By Mianjie Yu, Zhongmin Zhao, Zizhao Mo, Jianxiong Liao, Wenxuan Li, Yulin Qiao. https://arxiv.org/abs/2610.22015 https://arxiv.org/abs/2610.22015 Thu, 15 Oct 2026 14:56:00 GMT arXiv (systems & security) cs.DC Systems and databases arXiv (systems & security) cs.DC — Did: We propose RASA-DL, a resource-aware sub-network allocation framework that treats the maximum sub-network width assigned to each node as a deployment variable. Found: Under the Moderate and Severe regimes, the minimum-latency operating point improves accuracy over the largest globally feasible uniform deployment by 2.80 and 2.43 percentage points, respectively, at the same modeled round latency. By Zhuojun Tian, Ming Xiao, Mikael Skoglund. https://arxiv.org/abs/2610.21815 https://arxiv.org/abs/2610.21815 Thu, 15 Oct 2026 13:03:00 GMT arXiv (systems & security) cs.DC Systems and databases arXiv (systems & security) cs.DC — Did: We give an exact critical-path characterization of the makespan of micro-batch pipelines serving LLMs on heterogeneous GPU clusters, quantifying how the classic homogeneous GPipe formula errs. Found: These results quantify the GPipe overestimate, decompose the bias into a homogeneity term Sum_j (tau_max - tau_j) and a communication-ignoring term Sum_j d_j, and imply that balancing per-stage compute dominates communication reduction for steady-state throughput. By Zhu Wenbo. https://arxiv.org/abs/2610.21796 https://arxiv.org/abs/2610.21796 Thu, 15 Oct 2026 12:47:00 GMT arXiv (systems & security) cs.DB Systems and databases arXiv (systems & security) cs.DB — Did: We present an NL2SPARQL framework that places conceptual compositionality at the center of query interpretation and connects it to the relational composition required for binding and query construction. By Isabelle Mohr, John Dujany, Andre Freitas. https://dl.acm.org/doi/10.1145/3859286 https://dl.acm.org/doi/10.1145/3859286 Thu, 15 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Did: On the Internet, we use the Base64 format to represent binary data (e.g., images or cryptographic keys) as text. https://dl.acm.org/doi/10.1145/3857350 https://dl.acm.org/doi/10.1145/3857350 Thu, 15 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Found: Vision Transformers (ViTs) achieve state-of-the-art performance across various vision tasks but pose significant deployment challenges due to their high computational complexity and memory footprint. https://arxiv.org/abs/2610.21676 https://arxiv.org/abs/2610.21676 Thu, 15 Oct 2026 11:01:00 GMT arXiv (systems & security) cs.DB Systems and databases arXiv (systems & security) cs.DB — Did: We present ConStore, a content storage engine combining two complementary techniques to reduce end-to-end retrieval latency. Found: Separate experiments on 100M- and billion-vector indexes demonstrate SICP's applicability at larger scales. By Zhutao Zhuang, Nong Xiao, Zhiguang Chen, Yue Yu. https://arxiv.org/abs/2610.21500 https://arxiv.org/abs/2610.21500 Thu, 15 Oct 2026 08:37:00 GMT arXiv (systems & security) cs.DC Systems and databases arXiv (systems & security) cs.DC — Did: We study such a tier in a disaggregated deployment of GLM-5.3, an FP8 mixture-of-experts model with multi-head latent attention (MLA), on 8xB300 nodes running prefill context parallelism (PCP8). Found: The sharded store raises prefill throughput by 14%, and by 34% together with an upstream kernel fix. By Muhammad Ibrahim. https://arxiv.org/abs/2610.21264 https://arxiv.org/abs/2610.21264 Thu, 15 Oct 2026 05:00:00 GMT arXiv (systems & security) cs.DB Systems and databases arXiv (systems & security) cs.DB — Did: We present AdaBool, a SpBMM framework designed to reconcile this tension through an adaptive data representation. Found: We implement AdaBool end to end with low-level system optimizations, evaluate it on both production and public workloads, and show that it outperforms representative baselines by over an order of magnitude. By Xiangrui Yang, Yifan Lei, Jiawei Yang, Yekang Zhan, Yuxin Zheng, Lei Mao. https://arxiv.org/abs/2610.21249 https://arxiv.org/abs/2610.21249 Thu, 15 Oct 2026 04:44:00 GMT arXiv (systems & security) cs.OS Systems and databases arXiv (systems & security) cs.OS — Did: In this paper, we study LMCache through realworld multi-session workloads that span same/different context x same/different prompt, using over 100 stateless requests per workload. Found: Second, deduplication delivers a dramatic write reduction only for the exact-repeat workload: one set collapses from 1.62 GB to 11.75 MB of writes, while the other three sets remain in the 2-5 GB range. By Ying He, Dingsen Shi, Yanbo Dai, Donger Chen, Harry Fu, Chris Tsu. https://dl.acm.org/doi/10.1145/3848127 https://dl.acm.org/doi/10.1145/3848127 Wed, 14 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) https://dl.acm.org/doi/10.1145/3857806 https://dl.acm.org/doi/10.1145/3857806 Wed, 14 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Did: The generalized minimal residual (GMRES) algorithm is one of the key iterative algorithms for solving large-scale sparse linear systems, which finds extensive application in scientific and engineering computations. https://dl.acm.org/doi/10.1145/3858418 https://dl.acm.org/doi/10.1145/3858418 Wed, 14 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Did: Resistive random access memory (ReRAM) based analog processing-in-memory (APIM) architectures have been extensively explored for accelerating matrix multiplication in various artificial neural networks (ANNs). https://dl.acm.org/doi/10.1145/3858621 https://dl.acm.org/doi/10.1145/3858621 Wed, 14 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Did: The scalability of file system journaling faces significant challenges due to the increasing concurrency demands of modern multi-core CPUs and high throughput NVMe SSDs. https://dl.acm.org/doi/10.1145/3847118 https://dl.acm.org/doi/10.1145/3847118 Tue, 13 Oct 2026 12:00:00 GMT ACM TACO (Architecture and Code Optimization) Systems and databases ACM TACO (Architecture and Code Optimization) — Did: Power analysis is crucial in modern chip design flow.