SLShiyi Li

Shiyi Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Shiyi Li has a total of 36 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Wisdom From File Systems: Fast and Reliable Update for Graph-based Vector Search via Soft Insert

    by Yanqi Pan, Jiahao Chen, Hao Huang, Wen Xia, Xiangyu Zou, Shiyi Li, Chentao Wu, Jianzong Wang

  • Open Access

    Deduplication can Make File Systems Faster

    by Yanqi Pan, Wen Xia, Erci Xu, Hao Huang, Xiangyu Zou, Shiyi Li

    Emerging storage technologies, such as persistent memory and ultra-low latency SSD, enable the deduplication file system (DedupFS) to use non-cryptographic hash for fast fingerprinting. However, we find that the accelerated computation exposes another major performance penalty: the seemingly innocuous in-storage deduplication metadata maintenance incurs up to 38% overhead in the I/O path. We find the root cause is that deduplication incurs additional crash consistency overhead due to the maintenance of dedup-specific fingerprint-to-physical mappings. However, our insight, called deduplication-assisted crash consistency , shows that the deduplication mapping does not necessarily incur overhead, but rather can be exploited to improve performance. Our insight is based on two key observations: (1) The deduplication mapping can be merged with the file system logical-to-physical mapping, forming a logical-fingerprint-physical (LFP) mapping. Thus, we can persist deduplication metadata alongside file system metadata in a single I/O; (2) Furthermore, this mapping, containing fingerprint of the data block, can be leveraged to verify data integrity. Thus, we can achieve orderless data writes by verifying the integrity during recovery. We propose DarthFS to realize the efficiency of deduplication-assisted crash consistency. Using a series of techniques to manage data, metadata, and orderless I/O, DarthFS successfully leverages deduplication to improve file system I/O performance while retaining compatibility, effectiveness, and memory efficiency. Experiments across a range of synthetic and real-world workloads show that DarthFS outperforms existing file systems even under the workloads without duplicates, consistently outperforms existing DedupFSes, and can minimize deduplication-induced crash consistency overhead.

  • Open Access

    Towards Efficient Extendible Perfect Hashing for Hybrid PM-DRAM Memory

    by Hao Hu, Qi Chen, Yiming Yin, Xiangyu Zou, Ting Yao, Hongpeng Wang, Shiyi Li, Wen Xia

    Hashing is a widely used and efficient indexing mechanism for key-value storage. The emergence of persistent memory (PM) has further enhanced hash indexes by providing non-volatility and DRAM-like performance. However, current research on PM-based hash indexes primarily focuses on hardware-specific persistence optimizations or write performance optimization, while neglecting the critical aspect of read performance. Our study reveals that hash collisions significantly impair the read performance of hash indexes on PM. Thus, perfect hashing, which eliminates collisions, has the potential to improve read performance. Nevertheless, due to PM’s high access latency, the data movement overhead during hash table expansion and the random accesses introduced by perfect hashing itself present performance bottlenecks. In this article, we propose EEPH+, an efficient perfect hashing scheme for PM that effectively eliminates hash collisions to enhance read performance. EEPH+ employs three techniques to alleviate the aforementioned bottlenecks and improve performance: an extendible hashing technique to reduce the data movement overhead during hash table expansion, a hybrid PM–DRAM layout where the index structure resides in DRAM combined with a complement move algorithm to minimize the random accesses inherent to perfect hashing, and a batching and prefetching scheme to facilitate the parallel execution of hash computations and data accesses across multiple searches, thereby further boosting read performance. We compare EEPH+ with the state-of-the-art hash indexes on PM by conducting comprehensive experiments on several real-world read-intensive and read-skew workloads. The experimental results confirm the superiority of our EEPH+, which outperforms state-of-the-art hash indexes by up to 3.6× in read throughput and 3.14× in 99th percentile latency.


  • BIRD+: Design of a Lightweight Communication Compressor for Resource-Constrained Distribution Learning Platforms

    by Donglei Wu, Weihao Yang, Xiangyu Zou, Hao Feng, Dingwen Tao, Shiyi Li, Wen Xia, Binxing Fang

    The Top-K sparsification-based compression framework is extensively explored for reducing communication costs in distributed learning. However, we identified several issues with existing Top-K sparsification-based compression methods: ( i ) The limited compressibility of the Top-K parameter's indexes critically restricts the overall communication compression ratio; ( ii ) Several time-consuming compression operations significantly offset the benefits of communication compression; ( iii ) The use of error feedback techniques to maintain model quality results in a high memory footprint consumption. To solve these issues, we propose BIRD, a lightweight tensor-wise Bi-Random sampling strategy with an expectation invariance property. Specifically, BIRD applies a tensor-wise index sharing mechanism that reduces the index proportion by allowing multiple tensor elements to share a single index, thus improving the overall compression ratio. Additionally, BIRD replaces the time-consuming Top-K sorting with a faster Bi-Random sampling strategy based on the aforementioned index sharing mechanism, significantly reducing compression overheads; Moreover, BIRD establishes an expectation invariance property into the Bi-Random sampling to ensure an approximate unbiased representation for the $L_1$ -norm of the sampled tensors, effectively maintaining the model quality without incurring extra memory costs. We further optimize BIRD to BIRD+ by introducing the uniform distribution-based sampling and Gamma correction on the tensor-wise sampling process, achieving a more flexibly adjustment of the sparsity with better convergence performance. Experimental evaluations across multiple conventional distributed learning tasks demonstrate that compared to state-of-the-art approaches, BIRD+ achieves higher communication compression ratios up to 36.2 $\times$ and higher computation throughput up to 149.6 $\times$ while maintaining the model quality without incurring extra memory costs.


  • Open Access

    gPPM: A Generalized Matrix Operation and Parallel Algorithm to Accelerate the Encoding/Decoding Process of Erasure Codes

    by Shiyi Li, Qiang Cao, Shenggang Wan, Wen Xia, Changsheng Xie

    Erasure codes are widely deployed in modern storage systems, leading to frequent usage of their encoding/decoding operations. The encoding/decoding process for erasure codes is generally carried out using the parity-check matrix approach. However, this approach is serial and computationally expensive, mainly due to dealing with matrix operations, which results in low encoding/decoding performance. These drawbacks are particularly evident for newer erasure codes, including SD and LRC codes. To address these limitations, this article introduces the Partitioned and Parallel Matrix ( PPM ) algorithm. This algorithm partitions the parity-check matrix, parallelizes encoding/decoding operations, and optimizes calculation sequence to facilitate fast encoding/decoding of these codes. Furthermore, we present a generalized PPM ( gPPM ) algorithm that surpasses PPM in performance by employing fine-grained dynamic matrix calculation sequence selection. Unlike PPM, gPPM is also applicable to erasure codes such as RS code. Experimental results demonstrate that PPM improves the encoding/decoding speed of SD and LRC codes by up to 210.81%. Besides, gPPM achieves up to 102.41% improvement over PPM and 32.25% improvement over RS regarding encoding/decoding speed.


  • Open Access

    Smart-DNN+: A Memory-efficient Neural Networks Compression Framework for the Model Inference

    by Donglei Wu, Weihao Yang, Xiangyu Zou, Wen Xia, Shiyi Li, Zhenbo Hu, Weizhe Zhang, Binxing Fang

    Deep Neural Networks (DNNs) have achieved remarkable success in various real-world applications. However, running a Deep Neural Network (DNN) typically requires hundreds of megabytes of memory footprints, making it challenging to deploy on resource-constrained platforms such as mobile devices and IoT. Although mainstream DNNs compression techniques such as pruning, distillation, and quantization can reduce the memory overhead of model parameters during DNN inference, they suffer from three limitations: (i) low model compression ratio for the lightweight DNN structures with little redundancy, (ii) potential degradation in model inference accuracy, and (iii) inadequate memory compression ratio is attributable to ignoring the layering property of DNN inference. To address these issues, we propose a lightweight memory-efficient DNN inference framework called Smart-DNN+, which significantly reduces the memory costs of DNN inference without degrading the model quality. Specifically, ① Smart-DNN+ applies a layerwise binary-quantizer with a remapping mechanism to greatly reduce the model size by quantizing the typical floating-point DNN weights of 32-bit to the 1-bit signs layer by layer. To maintain model quality, ② Smart-DNN+ employs a bucket-encoder to keep the compressed quantization error by encoding the multiple similar floating-point residuals into the same integer bucket IDs. When running the compressed DNN in the user’s device, ③ Smart-DNN+ utilizes a partially decompressing strategy to greatly reduce the required memory overhead by first loading the compressed DNNs in memory and then dynamically decompressing the required materials for model inference layer by layer. Experimental results on popular DNNs and datasets demonstrate that Smart-DNN+ achieves lower 0.17%–0.92% memory costs at lower runtime overheads compared with the states of the art without degrading the inference accuracy. Moreover, Smart-DNN+ potentially reduces the inference runtime up to 2.04× that of conventional DNN inference workflow.


  • Top Cited Papers
    Open Access

    The Design of Fast and Lightweight Resemblance Detection for Efficient Post-Deduplication Delta Compression

    by Wen Xia, Lifeng Pu, Xiangyu Zou, Philip Shilane, Shiyi Li, Haijun Zhang, Xuan Wang

    Post-deduplication delta compression is a data reduction technique that calculates and stores the differences of very similar but non-duplicate chunks in storage systems, which is able to achieve a very high compression ratio. However, the low throughput of widely used resemblance detection approaches (e.g., N-Transform) usually becomes the bottleneck of delta compression systems due to introducing high computational overhead. Generally, this overhead mainly consists of two parts: ① calculating the rolling hash byte by byte across data chunks and ② applying multiple transforms on all of the calculated rolling hash values. In this article, we propose Odess, a fast and lightweight resemblance detection approach, that greatly reduces the computational overhead for resemblance detection while achieving high detection accuracy and a high compression ratio. Odess first utilizes a novel Subwindow-based Parallel Rolling (SWPR) hash method using Single Instruction Multiple Data [ 1 ] (SIMD) to accelerate calculation of rolling hashes (corresponding to the first part of the overhead). Odess then uses a novel Content-Defined Sampling method to generate a much smaller proxy hash set from the whole rolling hash set and quickly applies transforms on this small hash set for resemblance detection (corresponding to the second part of the overhead). Evaluation results show that during the stage of resemblance detection, the Odess approach is ∼31.4× and ∼7.9× faster than the state-of-the-art N-Transform and Finesse (a recent variant of N-Transform [ 39 ]), respectively. When considering an end-to-end data reduction storage system, the Odess-based system’s throughput is about 3.20× and 1.41× higher than the N-Transform- and Finesse-based systems’ throughput, respectively, while maintaining the high compression ratio of N-Transform and achieving ∼1.22× higher compression ratio over Finesse.


  • Open Access

    Use of fermented Chinese medicine residues as a feed additive and effects on growth performance, meat quality, and intestinal health of broilers

    by Xinhong Zhou, Shiyi Li, Yilong Jiang, Jicheng Deng, Chuanpeng Yang, Lijuan Kang, Huaidan Zhang, Xianxin Chen

    Introduction: The purpose of this research was to investigate how dietary supplementation with fermented herbal residues (FCMR) affected birds' development capacity, quality of meat, gut barrier, and cecum microbiota.Methods: 540 cyan-shank partridge birds aged 47 days were chosen and divided into two groups of six replicates each and 45 birds for each replicate. The control group (CON) received a basal diet, while the trial group decreased a basic diet containing 5% FCMR.Results and discussion: The findings revealed that the addition of FCMR decreased FCR and increased ADG in broilers (P < 0.05). Adding FCMR increased steaming loss in broiler chicken breasts (p < 0.05). Supplementation with FCMR significantly enhanced VH/CD and VH in the bird's intestine (jejunum, duodenum, and ileum) (p < 0.05). In addition, the addition of FCMR significantly down-regulated mRNA expression of INF-γ, IL-6, IL-1β, and TNF-α and up-regulated mRNA expression of ZO-1, Occludin, and Claudin (P < 0.05). Microbial 16S rDNA high-throughput sequencing study revealed that supplements with FCMR modified the cecum microbiota, and α-diversity analysis showed that supplementation with FCMR reduced the cecum bacterial abundance in broilers (P < 0.05). At the phylum level, the relative abundance of Spirochaetota increased considerably following FCMR supplementation (P < 0.05). The broiler cecum's close lot of Prevotellaceae_UCG-001 (P < 0.05), Desulfovibrio, Muribaculaceae, and Fusobacterium (p < 0.05) reduced when FCMR was supplemented. Supplementation with FCMR can promote growth capacity and maintain intestinal health in birds by enhancing gut barrier function and modulating the inflammatory response and microbial composition.