YZYue Zha

Yue Zha

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

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


Yue Zha has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Liquid Silicon: A Nonvolatile Fully Programmable Processing-in-Memory Processor With Monolithically Integrated ReRAM

    by Yue Zha, Etienne Nowak, Jing Li

    The slowdown of the CMOS technology scaling, and the trade-off between efficiency and flexibility have fueled the exploration into novel architectures with emerging postCMOS technology [e.g., resistive-RAM (RRAM)]. In this article, a nonvolatile fully programmable processing-in-memory (PIM) processor named liquid silicon is demonstrated, which combines the superior programmability of general-purpose computing devices [e.g., field-programmable gate array (FPGA)] and the high efficiency of domain-specific accelerators. Besides the general computing applications, liquid silicon is particularly well suited for artificial intelligence (AI)/machine learning and big data applications, which not only poses high computational/memory demand but also evolves rapidly. To fabricate the liquid silicon chip, the HfO₂ RRAM is monolithically integrated on top of the commercial 130 nm CMOS. Our measurement confirms that liquid silicon chip can operate reliably at a low voltage of 650 mV. It achieves 60.9 TOPS/W in performing neural network (NN) inferences, and 480 GOPS/W in performing content-based similarity search (a key big data application) at a nominal voltage supply of 1.2 V, showing 3x and 100x improvement over the state-of-the-art domain-specific CMOS-/RRAM-based accelerators. In addition, it outperforms the latest nonvolatile FPGA in energy efficiency by 3x in general computing applications.


  • MEG: A RISCV-Based System Simulation Infrastructure for Exploring Memory Optimization Using FPGAs and Hybrid Memory Cube

    by Jialiang Zhang, Yang Liu, Gaurav Jain, Yue Zha, Jonathan Ta, Jing Li

    Emerging 3D memory technologies, such as the Hybrid Memory Cube (HMC) and High Bandwidth Memory (HBM), provide increased bandwidth and massive memory-level parallelism. Efficiently integrating emerging memories into existing system pose new challenges and require detailed evaluation in a real computing environment. In this paper, we propose MEG, an open-source, configurable, cycle-exact, and RISC-V based full system simulation infrastructure using FPGA and HMC. MEG has three highly configurable design components: (i) a HMC adaptation module that not only enables communication between the HMC device and the processor cores but also can be extended to fit other memories (e.g., HBM, nonvolatile memory) with minimal effort, (ii) a reconfigurable memory controller along with its OS support that can be effectively leveraged by system designers to perform software-hardware co-optimization, and (iii) a performance monitor module that effectively improves the observability and debuggability of the system to guide performance optimization. We provide a prototype implementation of MEG on Xilinx VCU110 board and demonstrate its capability, fidelity, and flexibility on real-world benchmark applications. We hope that our open-source release of MEG fills a gap in the space of publicly-available FPGA-based full system simulation infrastructures specifically targeting memory system and inspires further collaborative software/hardware innovations.


  • Specialization: A New Path Towards Low Power

    by Yue Zha, Jing Li

    Today, energy-efficiency becomes the primary development goal across all computing platforms–from mobile devices to high-end servers. Despite that it has been steadily improved through technology and voltage scaling over the past decades, due to the fundamental physical limitation, it is changeling to retain the benefits using these traditional methods. We envision that specialization is a new path to extending the system scaling roadmap for low power. In this paper, we survey the recent progress on specialization, which can be broadly divided into two categories i.e., specialization in technology and specialization in architecture. For specialization in technology, we focus on emerging non-volatile memory (NVM) technologies (particularly, phase change memory) to provide a concrete review. For specialization in architecture, we first present an overview of NVM-based memory architectures, and then highlight the recent development of NVM-based non-Von Neumann architectures that provide processing-in-memory capability for various applications.

  • Open Access

    Liquid Silicon-Monona

    by Yue Zha, Jing Li

    With the recent trend of promoting Field-Programmable Gate Arrays (FPGAs) to first-class citizens in accelerating compute-intensive applications in networking, cloud services and artificial intelligence, FPGAs face two major challenges in sustaining competitive advantages in performance and energy efficiency for diverse cloud workloads: 1) limited configuration capability for supporting light-weight computations/on-chip data storage to accelerate emerging search-/data-intensive applications. 2) lack of architectural support to hide reconfiguration overhead for assisting virtualization in a cloud computing environment. In this paper, we propose a reconfigurable memory-oriented computing fabric, namely Liquid Silicon-Monona (L-Si), enabled by emerging nonvolatile memory technology i.e. RRAM, to address these two challenges. Specifically, L-Si addresses the first challenge by virtue of a new architecture comprising a 2D array of physically identical but functionally-configurable building blocks. It, for the first time, extends the configuration capabilities of existing FPGAs from computation to the whole spectrum ranging from computation to data storage. It allows users to better customize hardware by flexibly partitioning hardware resources between computation and memory, greatly benefiting emerging search- and data-intensive applications. To address the second challenge, L-Si provides scalable multi-context architectural support to minimize reconfiguration overhead for assisting virtualization. In addition, we provide compiler support to facilitate the programming of applications written in high-level programming languages (e.g. OpenCL) and frameworks (e.g. TensorFlow, MapReduce) while fully exploiting the unique architectural capability of L-Si. Our evaluation results show L-Si achieves 99.6% area reduction, 1.43× throughput improvement and 94.0% power reduction on search-intensive benchmarks, as compared with the FPGA baseline. For neural network benchmarks, on average, L-Si achieves 52.3× speedup, 113.9× energy reduction and 81% area reduction over the FPGA baseline. In addition, the multi-context architecture of L-Si reduces the context switching time to - 10ns, compared with an off-the-shelf FPGA (∼100ms), greatly facilitating virtualization.


  • RRAM-based reconfigurable in-memory computing architecture with hybrid routing

    by Yue Zha, Jing Li

    Recent advances in resistive random-access memory (RRAM) evoke great interests in exploring alternative architectures. One interesting work is a RRAM-based reconfigurable architecture that provides superior programmbility and blurs the boundary between computation and storage, but long-distance routing becomes a performance bottleneck. However, long-distance routing in FPGA is efficiently implemented, but its fine-grained routing structure results in a large routing overhead. In this work, we present a RRAM-based reconfigurable architecture that addresses the routing challenges using hybrid routing, i.e., local and global routing by taking the best advantages of both architectures (prior RRAM-based and FPGA). We also provide a complete CAD framework that exhibits high parallelism and good scalability. Experimental results show that our reconfigurable architecture outperforms both architectures. It achieves a 46.88% reduction in delay and improves the energy efficiency by 66.23% compared with the prior RRAM-based architecture with a slightly increased area overhead. While comparing with FPGA, it reduces the delay and the routing overhead by 36.00% and 50.20%, respectively. Additionally, our CAD framework achieves 5.39x speedup, compared with the prior framework.


  • CMA: A Reconfigurable Complex Matching Accelerator for Wire-Speed Network Intrusion Detection

    by Yue Zha, Jing Li

    The rapid growth in network bandwidth and the ever more sophisticated network attack techniques pose challenges to current network intrusion detection systems (NIDS). While software-based solutions are incapable of performing wire-speed network traffic monitoring, many hardware-based pattern matching solutions also suffer from capacity limitation and high power consumption. To effectively address these challenges, we propose a reconfigurable complex matching accelerator (CMA) enabled by the emerging nonvolatile memory technology (resistive random access memory) to speed up intrusion detection systems with better energy efficiency. Beyond common equality matching in current NIDS, CMA can be configured to provide a comprehensive set of arithmetic matching functions (e.g., less than), resulting in improved utilization and higher energy efficiency. We evaluate CMA using real-world network security benchmarks. On average, it achieves 84.9 percent area reduction, 97.3 percent energy consumption reduction, and 20 percent improvement in searching speed compared to the SRAM-based Ternary Content Addressable Memory (TCAM) design in state-of-the-art NIDS. It also outperforms emerging RRAM-based TCAM (2.5T1R) design in area, energy and search delay, on the set of evaluated workloads.


  • Open Access

    Challenges and Opportunities

    by Soroosh Khoram, Yue Zha, Jialiang Zhang, Jing Li

    The confluence of the recent advances in technology and the ever-growing demand for large-scale data analytics created a renewed interest in a decades-old concept, processing-in-memory (PIM). PIM, in general, may cover a very wide spectrum of compute capabilities embedded in close proximity to or even inside the memory array. In this paper, we present an initial taxonomy for dividing PIM into two broad categories: 1) Near-memory processing and 2) In-memory processing. This paper highlights some interesting work in each category and provides insights into the challenges and possible future directions.