In this paper, we propose a new dynamic reliability management technique for multicore processors using phase-driven Q-learning-based method. Our technique considers a wide range of long-term reliability issues and maximizes the throughput of the processor subject to the reliability constraint. We employ ON/OFF switching actions and dynamic voltage and frequency scaling as control knobs (i.e., working modes) to tune the state of cores of the processor. In order to achieve this, our technique detects program phases and adaptively determines the optimal working modes for each phase using the Q-learning-based method. By integrating the phase detection into the Q-learning-based management, our technique can provide efficient management for the programs with highly diverse phases. We also propose three additional modules to improve the management efficiency of our technique. In order to evaluate our technique, we use it to manage a 3-D CPU with high-diver programs. Several failure mechanisms are considered in this case study. Our proposed technique is compared with two existing Q-learning-based techniques. The experimental results demonstrate that when the number of phases is smaller than the number of working modes, our technique can achieve more than 1.36× improvement in performance with 60% memory space savings.

In this paper, we propose a phase-driven Q-learning based dynamic reliability management (DRM) technique for multi-core processors to solve DRM problems of maximizing the processor performance subject to a large class of reliability constraints by turning ON/OFF cores and dynamic voltage frequency scaling. Our technique utilizes the existing methods to detect program phases (i.e. [17]) and learns (rather than obtaining at the off-line stage) the optimal configuration of the multi-core processor for each phase. Our technique outperforms the existing learning-based DRM methods in managing programs with highly diverse phases. Our proposed technique is evaluated by solving a DRM problem in 3D CPUs of maximizing processor performance subject to the electromigration induced power delivery network reliability constraint. Compared to the latest Q-learning based DRM technique [11], our method can achieve more than 1.3x improvement in performance with 77% memory savings.

Vertically integrated circuits (3-D ICs) may revitalize Moore's law scaling which has slowed down in recent years. 3-D stacking is an emerging technology that stacks multiple dies vertically to achieve higher transistor density independent of device scaling. They provide high-density vertical interconnects, which can reduce interconnect power and delay. Moreover, 3-D ICs can integrate disparate circuit technologies into a single chip, thereby unlocking new system-on-chip architectures that do not exist in 2-D technology. While 3-D integration could bring new architectural opportunities and significant performance enhancement, new thermal, power delivery, signal integrity and reliability challenges emerge as power consumption grows, and device density increases. Moreover, the significant expansion of CPU design space in 3-D requires new architectural models and methodologies for design space exploration (DSE). New design tools and methods are required to address these 3-D-specific challenges. This keynote paper focuses on the state of the art, ongoing advances and future challenges of 3-D IC design tools and methods. The primary focus of this paper is TSV-based 3-D ICs, although we also discuss recent advances in monolithic 3-D ICs. The objective of this paper is to provide a unified perspective on the fundamental opportunities and challenges posed by 3-D ICs especially from the context of design tools and methods. We also discuss the methodology of co-design to address more complicated and interdependent design problems in 3-D IC, and conclude with a discussion of the remaining challenges and open problems that must be overcome to make 3-D IC technology commercially viable.

We propose efficient algorithms to construct a low-power clock tree for through-silicon-via (TSV)-based 3D-ICs. We use shutdown gates to save clock trees’ dynamic power, which selectively turn off certain clock tree branches to avoid unnecessary clock activities when the modules in these tree branches are inactive. While this clock gating technique has been extensively studied in 2D circuits, its application in 3D-ICs is unclear. In 3D-ICs, a shutdown gate is connected to a control signal unit through control TSVs, which may cause placement conflicts with existing clock TSVs in the layout due to TSV’s large physical dimension. We develop a two-phase clock tree synthesis design flow for 3D-ICs: (1) 3D abstract clock tree generation based on K-means clustering and (2) clock tree embedding with simultaneous shutdown gates’ insertion based on simulated annealing (SA) and a force-directed TSV placer. Experimental results indicate that (1) the K-means clustering heuristic significantly reduces the clock power by clustering modules with similar switching behavior and close proximity, and (2) the SA algorithm effectively inserts the shutdown gates to a 3D clock tree, while considering control TSV’s placement. Compared with previous 3D clock tree synthesis techniques, our K-means clustering-based approach achieves larger reduction in clock tree power consumption while ensuring zero clock skew.

Three-dimensional integration enables stacking DRAM on top of CPU, providing high bandwidth and short latency. However, non-uniform voltage fluctuation and local thermal hotspot in CPU layers are coupled into DRAM layers, causing a non-uniform bit-cell leakage (thereby bit flip) distribution. We propose a performance-power-resilience simulation framework to capture DRAM soft error in 3D multi-core CPU systems. A dynamic resilience management (DRM) scheme is investigated, which adaptively tunes CPU's operating points to adjust DRAM's voltage noise and thermal condition during runtime. The DRM uses dynamic frequency scaling to achieve a resilience borrow-in strategy, which effectively enhances DRAM's resilience without sacrificing performance.

In 3D-IC technology, electromigration (EM) degradation has become severe due to the high thermal-mechanical stress induced by the Through-Silicon-Vias (TSVs). However, little has been done on designing an EM-robust clock tree for 3D-ICs. In this paper, we propose a systematic EM-aware clock tree synthesis design flow, to enhance the 3D clock tree's EM reliability, with little interference to clock tree's performance metrics such as total wire length and clock skew. We develop a simple TSV's EM objective function based on multi-physics of the mass transportation equation, and validate it against the finite element method (FEM) simulation. Then we use this objective function to formulate a heuristic, based on integer linear programming (ILP), which places the clock TSVs such that the clock tree's EM reliability is maximized. Results show that the our heuristic is able to increase 3D clock's EM lifetime by more than 3.63x with little wire length overhead, while maintaining zero clock skew.

Through-silicon-via (TSV) offers vertical connections for 3-D ICs. Due to its large dimensions and nonideal etching process, TSVs layout needs to be carefully optimized to balance peak current density and delay for digital circuit. This brief investigates the TSVs tapering effect (which is an inevitable byproduct of deep reactive Ion etching-based manufacturing) and its impact on the TSVs electrical properties. We show that the current crowding effect is more severe in realistic tapered TSVs than ideal cylindrical TSVs. We propose a nonuniform current density model for tapered TSVs, which achieves considerable accuracy and speedup in estimating the current density distribution, when compared with the existing models developed for cylindrical TSVs. We apply our model to perform a detailed study on: 1) impact of TSVs tapering on peak current density and 2) wire sizing problem to minimize TSV-involved path delay under second-order delay model while keeping the peak current density within tolerable levels. A new dynamic programming-based heuristic is proposed to find the optimal wire configuration, which reduces both peak current density and delay, thereby improving the reliability and performance.

In this paper, we minimize 3D clock power using shutdown gates to selectively turn off unnecessary clock activities. In 3D-IC, shutdown signals require large-sized Through-Silicon-Vias(TSVs), so we propose a simulated annealing(SA) based algorithm along with a force-directed TSV placer to decide the selection of shutdown gates and the locations of TSVs under layout whitespace constraint. Furthermore, we recognize optimal power saving is achieved when the clock tree itself is designed simultaneously with the shutdown network. Experimental results show that our heuristic decreases the total clock power by more than 20% with less than 1.5% wirelength overhead while ensuring zero clock skew.