Junqiao Qiu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Junqiao Qiu has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Yunmo Zhang
City University of Hong Kong, Hong Kong, China
4 shared publications - Chun Jason Xue
Mohamed bin Zayed University of Artificial Intelligence
3 shared publications - Xizhe Yin
University of California, Riverside, Riverside, California, USA
3 shared publications - Hong Xu
The Chinese University of Hong Kong, Hong Kong, China
2 shared publications - Jiacheng Huang
City University of Hong Kong, Hong Kong, China
2 shared publications - Abhinav Vishnu
Advanced Micro Devices, Inc., United States of America
1 shared publication
Latest Publications
- 6 November 2026
- Published by Springer Nature
- p. 265-273
GCI: GPU-Centric Concurrent Incremental Analysis on Dynamic Graphs
- 6 November 2026
- Published by Springer Nature
- p. 265-273
GCI: GPU-Centric Concurrent Incremental Analysis on Dynamic Graphs
- Open Access
- 5 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 551-564
GAAF: Fast and Scalable Graph-based Vector Similarity Search with Any-Match Label Filtering
- 5 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 551-564
Open AccessGAAF: Fast and Scalable Graph-based Vector Similarity Search with Any-Match Label Filtering
- Open Access
- 16 March 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Vol. 10 (1) , 1-28
Visualizing Power Mode: The Impact of Battery-saving Indicators on User Behavior During Intensive Mobile Interactions
The ubiquitous mobile devices have made balancing battery efficiency and user experience (UX) a critical issue. While prior work has extensively researched battery-saving strategies by managing computational resources, the design of user interface (UI) that communicates battery-saving status with users remains underexplored. In this work, we investigate how different visual representations of battery-saving indicators influence user behavior during intensive mobile interactions. We designed three versions of UI indicators for battery saving, with varying levels of statistical details. Through a between-subjects user study with 36 participants completing a series of intensive tasks on a mobile phone with limited battery, we examined behavioral and perceptual patterns across study conditions. Our findings showed that real-time saving statistics improved battery-saving efficiency and mitigated negative user experience under moderate performance degradation. However, when performance drops sharply, the same indicators may bring frustration and lead users to quickly turn off the saver. Our post-study interviews further revealed the strategic choices made by participants to optimize task completion while saving batteries. With the lessons learned, we discuss the implications of designing visual indicators for battery-saving mode in different interaction scenarios, and propose future directions for building sustainable battery-saving interfaces.
- 16 March 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Vol. 10 (1) , 1-28
Open AccessVisualizing Power Mode: The Impact of Battery-saving Indicators on User Behavior During Intensive Mobile Interactions
The ubiquitous mobile devices have made balancing battery efficiency and user experience (UX) a critical issue. While prior work has extensively researched battery-saving strategies by managing computational resources, the design of user interface (UI) that communicates battery-saving status with users remains underexplored. In this work, we investigate how different visual representations of battery-saving indicators influence user behavior during intensive mobile interactions. We designed three versions of UI indicators for battery saving, with varying levels of statistical details. Through a between-subjects user study with 36 participants completing a series of intensive tasks on a mobile phone with limited battery, we examined behavioral and perceptual patterns across study conditions. Our findings showed that real-time saving statistics improved battery-saving efficiency and mitigated negative user experience under moderate performance degradation. However, when performance drops sharply, the same indicators may bring frustration and lead users to quickly turn off the saver. Our post-study interviews further revealed the strategic choices made by participants to optimize task completion while saving batteries. With the lessons learned, we discuss the implications of designing visual indicators for battery-saving mode in different interaction scenarios, and propose future directions for building sustainable battery-saving interfaces.
- Open Access
- 29 January 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 22 (3) , 1-18
C3: Finding Counting-related Crash Consistency Atomicity Bugs for Persistent Memory
Persistent Memory (PM) technologies provide fast, byte-addressable access to durable storage but face crash consistency challenges, motivating extensive work of testing and verification of PM programs. Central to PM testing tools is the specification of program properties for object persistence order and atomicity. Prior work on PM property inference has focused on ordering properties and offers limited support for reasoning about atomicity. This paper explores a class of important atomicity properties between the container-like arrays and their logical size variables, referred to as the counting correlation. These properties are prevalent in PM programs but are not adequately addressed by existing techniques. We introduce C3, a tool designed to infer these counting-related atomicity properties and detect related bugs in PM programs. Our approach begins by proposing invariants to capture the necessary behaviors of counting-correlated variables. We then utilize symbolic range analysis to extract PM program behaviors, and encode them into SMT constraints. These constraints are checked against the invariants to infer likely PM program properties. Our bug detection method is based on the existing output checking approach, but adapted to be aware of input properties to detect and confirm the atomicity bugs efficiently. Our evaluation shows that C3 has found 14 atomicity bugs (including 11 previously unknown ones) from real-world PM programs.
- 29 January 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 22 (3) , 1-18
Open AccessC3: Finding Counting-related Crash Consistency Atomicity Bugs for Persistent Memory
Persistent Memory (PM) technologies provide fast, byte-addressable access to durable storage but face crash consistency challenges, motivating extensive work of testing and verification of PM programs. Central to PM testing tools is the specification of program properties for object persistence order and atomicity. Prior work on PM property inference has focused on ordering properties and offers limited support for reasoning about atomicity. This paper explores a class of important atomicity properties between the container-like arrays and their logical size variables, referred to as the counting correlation. These properties are prevalent in PM programs but are not adequately addressed by existing techniques. We introduce C3, a tool designed to infer these counting-related atomicity properties and detect related bugs in PM programs. Our approach begins by proposing invariants to capture the necessary behaviors of counting-correlated variables. We then utilize symbolic range analysis to extract PM program behaviors, and encode them into SMT constraints. These constraints are checked against the invariants to infer likely PM program properties. Our bug detection method is based on the existing output checking approach, but adapted to be aware of input properties to detect and confirm the atomicity bugs efficiently. Our evaluation shows that C3 has found 14 atomicity bugs (including 11 previously unknown ones) from real-world PM programs.
- Open Access
- 8 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 564-579
PIE: Enabling Fast and Scalable Incremental Evolving Graph Analytics on Persistent Memory
- 8 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 564-579
Open AccessPIE: Enabling Fast and Scalable Incremental Evolving Graph Analytics on Persistent Memory
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Generative AI in Special Education: Teachers' Insights on Instructional Enrichment vs. Accommodations
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessGenerative AI in Special Education: Teachers' Insights on Instructional Enrichment vs. Accommodations
- Open Access
- 27 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 654-670
More Apps, Faster Hot-Launch on Mobile Devices via Fore/Background-aware GC-Swap Co-design
Faster app launching is crucial for the user experience on mobile devices. Apps launched from a background cached state, called hot-launching, have much better performance than apps launched from scratch. To increase the number of hot-launches, leading mobile vendors now cache more apps in the background by enabling swap. Recent work also proposed reducing the Java heap to increase the number of cached apps. However, this paper found that existing methods deteriorate app hot-launch performance while increasing the number of cached apps. To simultaneously improve the number of cached apps and hot-launch performance, this paper proposes Fleet, a foreground/background-aware GC-swap co-design framework. To enhance app-caching capacity, Fleet limits the tracing range of GC to background objects only, avoiding touching long-lifetime foreground objects. To improve hot-launch performance, Fleet identifies objects that will be accessed during the next hot-launch and uses runtime information to guide the swap scheme in the OS. In addition, Fleet aggregates small objects with similar access patterns into the same pages to improve swap efficiency. We implemented Fleet in AOSP and evaluated its performance with different types of apps. Experimental results show that Fleet achieves a 1.59× faster hot-launch time and caches 1.21× more apps than Android.
- 27 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 654-670
Open AccessMore Apps, Faster Hot-Launch on Mobile Devices via Fore/Background-aware GC-Swap Co-design
Faster app launching is crucial for the user experience on mobile devices. Apps launched from a background cached state, called hot-launching, have much better performance than apps launched from scratch. To increase the number of hot-launches, leading mobile vendors now cache more apps in the background by enabling swap. Recent work also proposed reducing the Java heap to increase the number of cached apps. However, this paper found that existing methods deteriorate app hot-launch performance while increasing the number of cached apps. To simultaneously improve the number of cached apps and hot-launch performance, this paper proposes Fleet, a foreground/background-aware GC-swap co-design framework. To enhance app-caching capacity, Fleet limits the tracing range of GC to background objects only, avoiding touching long-lifetime foreground objects. To improve hot-launch performance, Fleet identifies objects that will be accessed during the next hot-launch and uses runtime information to guide the swap scheme in the OS. In addition, Fleet aggregates small objects with similar access patterns into the same pages to improve swap efficiency. We implemented Fleet in AOSP and evaluated its performance with different types of apps. Experimental results show that Fleet achieves a 1.59× faster hot-launch time and caches 1.21× more apps than Android.
- 23 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 20 (2) , 1587-1593
Exploring Scalable Parallelization for Edit Distance-Based Motif Search
Motif Searching is an important problem that can reveal crucial information from biological data. Since the general motif searching is NP-hard and the volume of biological data is growing exponentially in recent years, there is a pressing need for developing time and space-efficient algorithms to find motifs. In this paper, we explore scalable parallelization for Edit Distance-Based Motif Search (EMS). We introduce two parallel designs, recursEMS which integrates the existing EMS solver into a parallel recursion tree running in multiple processes, and parEMS that presents a novel thread-based method which avoids the storage of redundant motif candidates. To make the parallel designs practical, we implement SPEMS , a S calability-sensitive P arallel solver for EMS. For any given biological dataset and search instance, SPEMS can provide an EMS parallelization towards the optimal performance, or a sub-optimal performance but being more space efficient. Evaluations on two real-world DNA dataset TRANSFAC and ChIP-seq show that SPEMS can obtain 10× geometric mean speedup over the state-of-the-art at the expense of no less than 74.7% memory overheads, or provide 2.2× geometric mean speedup with the possibility of consuming less memory, when running on a 48-core machine.
- 23 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 20 (2) , 1587-1593
Exploring Scalable Parallelization for Edit Distance-Based Motif Search
Motif Searching is an important problem that can reveal crucial information from biological data. Since the general motif searching is NP-hard and the volume of biological data is growing exponentially in recent years, there is a pressing need for developing time and space-efficient algorithms to find motifs. In this paper, we explore scalable parallelization for Edit Distance-Based Motif Search (EMS). We introduce two parallel designs, recursEMS which integrates the existing EMS solver into a parallel recursion tree running in multiple processes, and parEMS that presents a novel thread-based method which avoids the storage of redundant motif candidates. To make the parallel designs practical, we implement SPEMS , a S calability-sensitive P arallel solver for EMS. For any given biological dataset and search instance, SPEMS can provide an EMS parallelization towards the optimal performance, or a sub-optimal performance but being more space efficient. Evaluations on two real-world DNA dataset TRANSFAC and ChIP-seq show that SPEMS can obtain 10× geometric mean speedup over the state-of-the-art at the expense of no less than 74.7% memory overheads, or provide 2.2× geometric mean speedup with the possibility of consuming less memory, when running on a 48-core machine.
- Open Access
- 14 June 2017
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Enabling scalability-sensitive speculative parallelization for FSM computations
Finite state machines (FSMs) are the backbone of many applications, but are difficult to parallelize due to their inherent dependencies. Speculative FSM parallelization has shown promise on multicore machines with up to eight cores. However, as hardware parallelism grows (e.g., Xeon Phi has up to 288 logical cores), a fundamental question raises: How does the speculative FSM parallelization scale as the number of cores increases? Without answering this question, existing methods for speculative FSM parallelization simply choose to use all available cores, which might not only waste computing resources, but also result in suboptimal performance. In this work, we conduct a systematic scalability analysis for speculative FSM parallelization. Unlike many other parallelizations which can be modeled by the classic Amdahl's law or its simple extensions, speculative FSM parallelization scales unconventionally due to the non-deterministic nature of speculation and the cost variations of misspeculation. To address these challenges, this work introduces a spectrum of scalability models that are customized to the properties of specific FSMs and the underlying architecture. The models, for the first time, precisely capture the scalability of speculative parallelization for different FSM computations, and clearly show the existence of a "sweet spot" in terms of the number of cores employed by the speculative FSM parallelization to achieve the optimal performance. To make the scalability models practical, we develop S3, a scalability-sensitive speculation framework for FSM parallelization. For any given FSM, S3 can automatically characterize its properties and analyze its scalability, hence guide speculative parallelization towards the optimal performance and more efficient use of computing resources. Evaluations on different FSMs and architectures confirm the accuracy of the proposed models and show that S3 achieves significant speedup (up to 5X) and energy savings (up to 77%).
- 14 June 2017
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Open AccessEnabling scalability-sensitive speculative parallelization for FSM computations
Finite state machines (FSMs) are the backbone of many applications, but are difficult to parallelize due to their inherent dependencies. Speculative FSM parallelization has shown promise on multicore machines with up to eight cores. However, as hardware parallelism grows (e.g., Xeon Phi has up to 288 logical cores), a fundamental question raises: How does the speculative FSM parallelization scale as the number of cores increases? Without answering this question, existing methods for speculative FSM parallelization simply choose to use all available cores, which might not only waste computing resources, but also result in suboptimal performance. In this work, we conduct a systematic scalability analysis for speculative FSM parallelization. Unlike many other parallelizations which can be modeled by the classic Amdahl's law or its simple extensions, speculative FSM parallelization scales unconventionally due to the non-deterministic nature of speculation and the cost variations of misspeculation. To address these challenges, this work introduces a spectrum of scalability models that are customized to the properties of specific FSMs and the underlying architecture. The models, for the first time, precisely capture the scalability of speculative parallelization for different FSM computations, and clearly show the existence of a "sweet spot" in terms of the number of cores employed by the speculative FSM parallelization to achieve the optimal performance. To make the scalability models practical, we develop S3, a scalability-sensitive speculation framework for FSM parallelization. For any given FSM, S3 can automatically characterize its properties and analyze its scalability, hence guide speculative parallelization towards the optimal performance and more efficient use of computing resources. Evaluations on different FSMs and architectures confirm the accuracy of the proposed models and show that S3 achieves significant speedup (up to 5X) and energy savings (up to 77%).