YSYunpeng Song

Yunpeng Song

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yunpeng Song has a total of 21 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Waltz: Temperature-Aware Cooperative Compression for High-Performance Compression-Based CSDs

    by Dingcui Yu, Yunpeng Song, Yiyang Huang, Yumiao Zhao, Yina Lv, Chundong Wang, Youtao Zhang, Liang Shi

    Data compression is widely adopted for modern solid-state drives (SSDs) to mitigate SSD lifetime issues. Researchers have proposed compression schemes at different system layers, including device-side solutions like CCSDs ( c ompression-based c omputational S SD s) and compression supported by host-side, like F2FS (flash-friendly file system). As storage performance increases, the surging power density makes thermal management a critical challenge. We conduct quantitative studies to understand how host-side and device-side compression schemes affect the temperature and performance of SSD-based storage systems. From our experiments, device-side compression, facilitated by a hardware compression engine, can raise the temperature of CCSDs to intolerable levels, resulting in throttling and service shutdown. In contrast, host-side compression causes software-stack overhead, which often results in large performance degradation and resource consumption. To ensure efficient data compression with high performance and better temperature control, we propose Waltz, a temperature-aware cooperative compression method that schedules (de)compression tasks at the host and device sides by monitoring device temperature. Furthermore, we introduce two variants (Waltzs and Waltzp) for space and performance optimization, respectively. Waltz is implemented within F2FS, achieving high performance while extending SSD lifetime and preventing overheating-induced in-flight shutdowns.

  • Prophet: SSD Failure Analysis and Prediction Guided by Flash Reliability Characteristics in Data Centers

    by Yunpeng Song, Yujiong Liang, Jialin Liu, Liang Shi

    Solid-state drives (SSDs) are massively deployed in various fields, especially in data centers, for their excellent cost-effectiveness. However, SSDs may fail due to their imperfect manufacturing processes, resulting in system-level failures and even downtime in data centers. This makes SSD failure prediction critical. Current studies focus on dealing with data missing, numerical normalization, and other statistical issues in using machine learning methods, but the consideration of the reliability characteristics of the underlying flash media of SSDs and the timeliness (time duration between predicted failure and real failure) of SSD failure prediction result is missing. In this work, we study the failure characteristics of over 200,000 drives from industry data centers over a 4-year period, as well as daily data. The relationship between SSD attribute values and failures is first investigated. Then, we analyzed the SSD failure characteristics from several aspects (causes, differences between failures, and timeliness of prediction results) relying on flash reliability characteristics. Based on these, a novel SSD failure prediction method (Prophet) is proposed. Specifically, Prophet contains the following two components. First, to cope with the differences between failures, a diff-state method is proposed for differential machine learning modeling of SSDs in different “States”. We define the “State” of an SSD, which represents the range of values in which the SSD currently lies in terms of some key attributes. Through flash reliability characteristics, we distinguish between different failures before training the model to obtain accurate predictions of different failure behaviors. Second, a recovery period method is proposed to enhance the timeliness of SSD failure prediction result by designing the sample selection method. The enhanced timeliness can be utilized by operations personnel to handle failed SSDs, such as replacement and repair. The evaluation results of the real dataset show that the predictive ability of Prophet is improved amazingly, realizing a high recall and low false-positive rates while providing sufficient response time for the processing of failed SSDs.


  • Revisiting Multiple ECC on High-Density NAND Flash memory

    by Yunpeng Song, Yina Lv, Wentong Li, Jialin Liu, Liang Shi

    Three-dimensional nand flash memory using the advanced multibit-per-cell technique is widely adopted due to its high density. However, it faces the problem of deteriorating read performance and energy consumption due to decreased reliability. Low-density parity-check code (LDPC) is typically adopted as an error correction code (ECC) to encode data and provide fault tolerance. To reduce the cost, LDPC with a high code rate is always adopted. However, LDPC will lead to read retry operations when the accessed data are not successfully decoded, and such retry-induced performance degradation is serious, especially for modern high-density flash memory. In this work, a reliability-aware differential ECC (READECC) approach is proposed to reduce redundancy protection and storage cost of LDPC with a low code rate and optimize the read performance. The basic idea is to adopt LDPC with a suitable code rate considering both data access characteristics and flash reliability characteristics. First, hot reads are identified based on the frequency of being accessed. Second, based on the reliability variation characteristics, the life of flash memory is divided into three reliability periods. As the reliability period shifts, the code rate of the LDPC adjusts adaptively to minimize redundancy protection. Third, an adaptive-sized logical page approach is further proposed to support LDPC with strong error correction capability (a low code rate) with a low storage cost. Through careful design and evaluation on 3-D triple-level-cell nand flash memory, READECC achieves encouraging optimizations with a negligible cost.


  • Temperature-Aware Differential Programming for Performance and Energy Optimization on 3D NAND High-Density Flash Memory

    by Yunpeng Song, Dingcui Yu, Zhonghuan Yan, Yanyun Wang, Liang Shi

    3D NAND high-density flash memory is widely used in edge computing, IoT, and automotive applications due to its high performance, low latency, and low storage cost characteristics. These scenarios require operation in extreme temperature environments, with cross-temperature read/write occurring frequently. However, cross-temperature affects programming reliability, leading to high raw bit error rates (RBER), which degrades read performance and increases energy consumption. In this paper, we propose a novel temperature-aware differential programming (TADP) scheme to optimize read performance and energy consumption under cross-temperature read/write. Specifically, first, a temperature-aware compensatory programming scheme is proposed to reduce the cross-temperature-induced degradation of RBER. Second, a layer variation-aware compensatory programming scheme is proposed to reduce the compensatory programming latency. Finally, a degraded programming scheme is proposed to enhance the temperature toughness of poorly temperature-tough word-lines by using them as MLC. Evaluated on 233-layer 3D triple-level-cell (TLC) NAND flash, TADP achieved encouraging optimizations in programming reliability, energy consumption, and read performance with minimal capacity loss.


  • Open Access

    Access Characteristic-Guided Remote Swapping Across Mobile Devices

    by Wentong Li, Yina Lv, Longfei Luo, Yunpeng Song, Liang Shi

    Memory swapping ensures smooth application switching for mobile systems by caching applications in the background. To further play the role of memory swapping, remote swapping across mobile devices has been widely studied, which caches applications to nearby remote devices by remote paging. However, due to the massive remote I/Os and unguaranteed swap throughput, the current remote swapping is limited with an unsatisfactory user experience, especially under variable network conditions. This paper first studies the access characteristics of applications and clarifies the impact of various network traffic on remote swapping. Motivated by these, an efficient access characteristic-guided remote swapping framework (ACR-Swap + ) is proposed to optimize remote swapping across mobile devices with resilient remote paging. ACR-Swap + first performs selective remote paging based on the swap-in frequency of different processes and then prefetches data across devices based on the process running states. Finally, it conducts hierarchical remote paging to avoid the impact of network traffic on remote swapping. Evaluations on Google Pixel 6 show that ACR-Swap + reduces the application switching latency by 21.6% and achieves a negligible performance fluctuation under various network traffic compared to the state of the art.


  • Open Access

    ElasticZRAM: Revisiting ZRAM for Swapping on Mobile Devices

    by Wentong Li, Dingcui Yu, Yunpeng Song, Longfei Luo, Liang Shi

    Modern mobile devices adopt two-level memory swapping consisting of ZRAM and storage devices to relieve memory pressure. In the swap subsystem, ZRAM can improve application responsiveness and reduce write traffic to storage devices while consuming physical memory and additional CPU cycles. To better utilize ZRAM and improve system performance, we propose ElasticZRAM, an elastic ZRAM to redesign the traditional memory swapping with full awareness of the characteristics of applications and NAND flash-based storage devices on mobile devices. Experimental results on Google Pixel 6 demonstrate that ElasticZRAM improves application response time by up to 24.8% with negligible overhead compared with state-of-the-arts.


  • Revisiting TRIM on High-Density Flash-Based Hybrid Storage Systems

    by Longfei Luo, Dingcui Yu, Hang Li, Yunpeng Song, Yina Lv, Edwin H.-M. Sha, Liang Shi

    Hybrid solid state drives (SSDs) that integrate high-performance (HP) and large-capacity (LC) flash are widely used due to their cost-effectiveness. The TRIM command, which is a popular command in normal SSDs to improve performance and endurance, is also recommended in hybrid SSDs. However, employing TRIM on hybrid SSDs as on normal SSDs will induce performance loss and suboptimal endurance due to the different characteristics of flash in hybrid SSDs. To solve the problem, this article first explores the critical factors of issuing TRIM commands to different flash. Then, this article proposed a differential TRIM method (dTRIM), which suggests performing early TRIM on HP flash and lazy TRIM on high-capacity flash. Specifically, early TRIM will minimize garbage collection costs while lazy TRIM tries to avoid conflicting user requests. Experimental results demonstrate that dTRIM can significantly improve the performance and endurance of hybrid SSDs compared with the state-of-the-arts.


  • Adaptive Differential Wearing for Read Performance Optimization on High-Density nand Flash Memory

    by Yunpeng Song, Yina Lv, Liang Shi

    With cost reduction and density optimization, high-density NAND flash memory has been widely deployed in data centers and consumer devices. However, this trend has significantly degraded the read performance and lifetime of high-density NAND flash memory during the last decade. Previous works proposed to optimize flash lifetime with wear leveling (WL) and optimize read performance with reliability improvement. Although WL can improve flash lifetime, it leads to the reliability of all blocks in 3-D NAND flash decreasing simultaneously. The reliability and read performance will be degraded with flash wearing. To solve this problem, an adaptive differential wearing (ADWR) scheme is proposed to optimize the read performance and lifetime in this work. The basic idea of ADWR is to determine the size of the high-reliability area to serve hot reads based on workload characteristics. Specifically, first, a differential wearing scheme is proposed to construct different reliability areas based on the characteristics of the data. Second, a lifetime model is constructed for the ADWR to clarify the lifetime impact. Based on this, a lifetime optimization scheme is proposed to improve the flash lifetime. Finally, a differential refresh scheme is proposed to reduce the impact of read disturbance on read performance. The experiments on real-life workloads show that ADWR achieves encouraging read performance optimization with negligible impacts on the lifetime of 3-D TLC NAND flash memory.


  • Open Access

    When F2FS Meets Compression-Based SSD!

    by Yunpeng Song, Yiyang Huang, Yina Lv, Yi Zhang, Liang Shi

    Compression-based schemes have been widely studied to improve the lifetime and performance of solid-state drives (SSDs). Recently, the most popular flash-friendly file system (F2FS) started supporting compression to maximize the lifetime of NAND flash-based storage. Also, compression-based computational SSDs (CSDs) are developed due to their high performance, transparency, and easy adoption. This paper will first study the compression of F2FS and CSD to understand their features. Then, cooperative compression (COCO) is proposed to optimize performance and power consumption based on the combination of F2FS and CSD. Experiments on real devices show that COCO has encouraged optimization.