Energy efficiency has become a critical design metric for high-performance systems. Various power management techniques have been proposed for the processor cores such as dynamic voltage and frequency scaling (DVFS), whereas few solutions consider the power losses suffered on the power delivery system (PDS), despite the fact that they have a significant impact on the overall energy efficiency of the system. With the explosive growth of system complexity and highly dynamic workloads variations, it is also challenging to find the optimal power management policies which can effectively match the power delivery with the power consumption. In addition, process variations (PVs) add heterogeneity to systems and make traditional power management methods less effective. To tackle the above problems, we propose a reinforcement-learning-based Chip-Specific Power co-Management (CSPM) scheme for PV-aware manycore systems. Both PDS and processor cores are jointly adjusted by distributed agents with modular Q-learning to improve the overall energy efficiency of the system. System characteristics are naturally included in the learning process to obtain chip-specific policies. Experimental results show that when applied to PV-aware manycore systems with a hybrid PDS constructed by both on- and off-chip voltage regulators, the proposed method achieves a 60.1% reduction of the overall energy delay product (EDP) of the system, on average, compared to a traditional DVFS approach.

Minimizing energy consumption while satisfying the user-specified performance requirement is a primary design objective for mobile devices. To achieve this, off-the-shelf mobile devices are usually equipped with dynamic voltage and frequency scaling (DVFS)-enabled processors to tradeoff performance for energy reduction through online adjustment of operating points. Recently, reinforcement learning algorithms have been widely studied and show great potential for runtime DVFS control because of their adaptability to the changing environment. However, the rapid evolution of hardware and the ever-growing diversity of mobile applications increase the system complexity dramatically and make it hard for the learning agent to quickly obtain an efficient power management policy. To address this challenge, we propose a decentralized collaborative Q-learning (DCQL)-based approach in this article to solve the DVFS control problem of multicore mobile processors. By exchanging learning experiences and knowledge among multiple devices, DCQL increases the convergence rate of the learning algorithm and improves the quality of the derived power management policy. On each device, the 4-phase action selection strategy is applied for efficient exploration of the action space defined by various power settings. Experimental results on realistic applications show that DCQL can achieve 3.4%-8.0% energy reduction over various existing approaches while providing an average of 3.4× speedup over the state-of-the-art individual learning algorithm. We also show that the proposed approach can scale well with the number of cores through clustered DVFS control.

Rapidly evolving embedded applications continuously demand more functionalities and better performance under tight energy and thermal budgets, and maintaining high energy efficiency has become a significant design challenge for mobile devices. Learning-based methods are adaptive to dynamic conditions and show great potential for runtime power management. However, with the ever-increasing complexity of both hardware and software, it is a challenging issue for a learning agent to explore the state-action space sufficiently and quickly find an efficient management policy. In this paper, we propose a reinforcement learning-based multi-device collaborative power management approach to address this issue. Multiple devices with different runtime conditions can acquire related knowledge during the learning process. Efficient knowledge sharing among these devices can potentially accelerate the learning process and improve the quality of the learned policies. We integrate the proposed method with dynamic voltage and frequency scaling on the multicore processors in mobile devices. Experimental results on realistic applications show that the collaborative power management can achieve up to a $7 \times$ speedup and 10% energy reduction compared with state-of-the-art learning-based approaches.

Rapidly evolving embedded applications continuously demand more functionality and better performance under tight energy and thermal budgets, and maintaining high energy efficiency has become a significant design challenge for mobile devices. Although learning-based runtime power management can adapt to dynamic conditions, it is a challenging issue to quickly find an efficient management policy under ever-increasing hardware and software complexity. In this work, we propose a multi-device collaborative power management approach to address this issue. The collaborative power management periodically shares knowledge among multiple devices to accelerate the learning process and improve the quality of learned policies. We integrate the proposed method with dynamic voltage and frequency scaling (DVFS) on the multicore processors in mobile devices. Experimental results on realistic applications show that the collaborative power management can achieve on average 8x speedup and 10% energy saving compared with state-of-the-art learning-based approaches.

Maintaining high energy efficiency has become a critical design issue for high-performance systems. Many power management techniques have been proposed for the processor cores such as dynamic voltage and frequency scaling (DVFS). However, very few solutions consider the power losses suffered on the power delivery system (PDS), despite the fact that they have a significant impact on the system overall energy efficiency. With the explosive growth of system complexity and highly dynamic workloads variations, it is challenging to find the optimal power management policies which can effectively match the power delivery with the power consumption. To tackle the above problems, we propose a reinforcement learning-based power management scheme for manycore systems to jointly monitor and adjust both the PDS and the processor cores aiming to improve system overall energy efficiency. The learning agents distributed across power domains not only manage the power states of processor cores but also control the on/off states of on-chip VRs to proactively adapt to the workload variations. Experimental results with realistic applications show that when the proposed approach is applied to a large-scale system with a hybrid PDS, it lowers the system overall energy-delay-product (EDP) by 41% than a traditional monolithic DVFS approach with a bulky off-chip VR.

Battery-powered mobile devices have limited energy capacity, urging the development of efficient power management approaches. Reinforcement learning (RL) algorithms are adaptive to the changing environment and have been widely used for runtime power management. Recently, collaborative RL-based approaches have been explored to accelerate the learning process. Nonetheless, existing works usually require a cloud service provider to achieve centralized multi-device knowledge sharing, which suffers from the single point of failure and does not guarantee users' privacy. To address this issue, we propose a decentralized multi-device collaborative power management approach in this work, where devices directly share their knowledge with their trusted neighbors. Experimental results show that the proposed method can achieve an up to 21% energy reduction with a 4× speedup over the individual learning-based approach.

The power delivery system (PDS), which plays a crucial role in guaranteeing the proper functionality of computing systems, has been a serious constraint on performance due to its significant power loss, especially for high-performance many-core processors. As the PDS design is usually optimized to provide power to the target chip at its best performance level, the energy efficiency can be notably degraded when the workload of the processors is highly dynamic. Therefore, dynamically adjusting the PDS to adapt to the run-time chip workloads are expected to be beneficial. In addition, the introduction of on-chip voltage regulators (VRs) has also significantly broadened the design space of PDS. In this paper, we present a workload-aware quantized power management scheme to dynamically manage the PDS in order to improve system energy efficiency. VRs at different stages are scheduled as online or offline based on chip workload estimation and prediction. The simulation results show that with the proposed scheme, a hybrid PDS with on/off-chip VRs can achieve 74.6% overall efficiency on average, 12.7% higher than a conventional PDS with one off-chip VR. The proposed scheme also shows its potential advantage in improving system energy efficiency with large-scale many-core processors and with more advanced processor technology nodes.

Energy-efficiency is becoming increasingly important to modern computing systems with multi-/many-core architectures. Dynamic Voltage and Frequency Scaling (DVFS), as an effective low-power technique, has been widely applied to improve energy-efficiency in commercial multi-core systems. However, due to the large number of cores and growing complexity of emerging applications, it is difficult to efficiently find a globally optimized voltage/frequency assignment at runtime. In order to improve the energy-efficiency for the overall multicore system, we propose an online DVFS control strategy based on core-level Modular Reinforcement Learning (MRL) to adaptively select appropriate operating frequencies for each individual core. Instead of focusing solely on the local core conditions, MRL is able to make comprehensive decisions by considering the running-states of multiple cores without incurring exponential memory cost which is necessary in traditional Monolithic Reinforcement Learning (RL). Experimental results on various realistic applications and different system scales show that the proposed approach improves up to 28% energy-efficiency compared to the recent individual-RL approach.

The memory wall problem is due to the imbalanced developments and separation of processors and memories. It is becoming acute as more and more processor cores are integrated into a single chip and demand higher memory bandwidth through limited chip pins. Optical memory interconnection network (OMIN) promises high bandwidth, bandwidth density, and energy efficiency, and can potentially alleviate the memory wall problem. In this paper, we propose an optical inter/intra-chip processor-memory communication architecture, called MOCA. Experimental results and analysis show that MOCA can significantly improve system performance and energy efficiency. For example, comparing to Hybrid Memory Cube (HMC), MOCA can speedup application execution time by 2.6x, reduce communication latency by 75%, and improve energy efficiency by 3.4x for 256-core processors in 7 nm technology.

The power delivery system (PDS) plays a crucial role of guaranteeing the proper functionality of many-core processors. However, as PDS is usually optimized to provide power to the target chip at its best performance level, its energy efficiency can be seriously degraded under highly dynamic workloads, making it a major source of system power losses. On-chip voltage regulators (VR), which are able to achieve fast and fine-grained power control, have been popular choices for PDS implementation and provided design opportunities for improving system energy efficiency. In this paper, we propose the adaptive Quantized Power Management (QPM) scheme to dynamically adjust the PDS with both on-chip and off-chip VRs based on run-time workloads. Experimental results on different applications show that QPM applied on a hybrid PDS with both on/off-chip voltage regulators (VR) achieves 74.1% average overall energy efficiency, 12.3% higher than the conventional PDS with single off-chip VR.

The design of power delivery system plays a crucial role in guaranteeing the proper functionality of many-core processor systems. The power loss suffered on power delivery has become a salient part of total power consumption, and the energy efficiency of a highly dynamic system has been significantly challenged. Being able to achieve a fast response time and multiple voltage domain control, on-chip voltage regulators (VRs) have become popular choices to enable fine-grain power management, which also enlarge the design space of power delivery systems. This paper analytically studies different power delivery system paradigms and power management schemes in terms of energy efficiency, area overhead, and power pin occupation. The analysis shows that compared to the conventional paradigm with off-chip VRs, hybrid paradigms with both on-chip and off-chip VRs are able to maintain high efficiency in a larger range of workloads, though they suffer from low efficiency at light workload. Employed with the quantized power management scheme, the hybrid paradigm can improve the system energy efficiency at light workload by a maximum of 136% compared to the traditional load balanced scheme. Besides this, the in-package (iP) hybrid paradigm further shows its advantage in reducing the physical overheads. The results reveal that at 120 W workload, it occupies only a 10.94% total footprint area or 39.07% power pins of that of the off-chip paradigm. We conclude that the iP hybrid paradigm achieves the best tradeoffs between efficiency, physical overhead, and realization of fine-grain power management.

The number of chip pins is limited due to the cost and reliability issues of sophisticated packages, and it is predicted that the chip pin count will be overstretched to satisfy the requirements of both power delivery and memory access. The gap between the achievable pin count and the demand will increase as the technology scales, due to the increasing computation resources and supply current. Pin reduction techniques are thus required for continued computing performance growth. In this article, we propose a chip pin constraint alleviation strategy, through on/off-chip power delivery system co-design, to effectively reduce the demand for power pins. An analytical model of a power delivery system, consisting of on/off-chip regulators and a power delivery network, is proposed to evaluate the influence of regulator design and package conduction loss. By combining this model with a multi-core processor model of performance and memory bandwidth requirements, we characterize the entire multi-core processor system to investigate the relationship between the chip pin constraint and performance in multi-core processor scaling and the effectiveness of our strategy. Experiments show that with the conventional power delivery system design, the chip pin constraint severely limits the performance growth as the technology scales. Using the on/off-chip power delivery system co-design, our strategy achieves a significant pin count reduction, for example, 31.3% at the 8nm technology node, compared to the conventional design with the same chip performance, while, provided with the same chip pin count, it is able to improve, by 35.0%, the chip performance at 8nm compared to the conventional design. For real applications of different parallelism, our strategy outperforms its counterpart, with a 23.7% performance improvement on average at the 8nm technology node.

Recently, interchip/intrachip optical interconnection networks have been proposed for ultrahigh-bandwidth and low-latency communications. These networks employ the microresonators (MRs) to modulate, direct, or detect the optical signal. However, utilized MRs suffer from intrinsic crosstalk noise and signal power loss, degrading the network efficiency via the signal-to-noise ratio (SNR). The amount of crosstalk noise and signal power loss may differ from network to network. Hence, there exists a need to systematically analyze the effect of the crosstalk noise and the power loss issues. In this paper, we have developed the analytical models considering both coherent and incoherent crosstalk for both the interchip and intrachip optical networks. The interchip/intrachip optical interconnection networks—the$\text{I}^{2}$CON—are analyzed as a case study. The quantitative results on the individual networks have demonstrated that the architectural design determines the impact of crosstalk on the SNR. We have also demonstrated that the optical interconnection networks with interchip/intrachip interconnects result in better bit error rate (BER) compared with that of only intrachip interconnect. Our analyses of the worst case can be utilized as a platform to compare the realistic performance among different optical interconnection networks via the degradation of SNR/BER and data bandwidth.

Power gating (PG) is one of the most effective techniques to reduce the leakage power in multiprocessor system-on-chips (MPSoCs). However, the power-mode transition during the PG period of an individual processing unit (PU) will introduce serious power/ground (P/G) noise to the neighboring PUs. As technology scales, the P/G noise problem becomes a severe reliability threat to MPSoCs. At the same time, the increasing manufacturing process variations (PVs) also bring uncertainties to the P/G noise problem and make it difficult to predict and mitigate. To tackle this problem, in this paper, we analyze the PG-induced P/G noise in the presence of PVs and propose a hardware–software collaborated runtime technique to adaptively protect PUs from P/G noise. Sensor network-on-chip is used to gather noise information and coordinate different system components. An online PV-aware algorithm is developed to effectively decide the noise impact range and arrange protections for affected PUs based on the collected noise information. We evaluate the proposed technique through cycle-level Monte Carlo simulations of NoC-based MPSoCs in different scales. The experimental results on various realistic applications show that our technique could achieve comparable reliability to the most reliable static technique while improve on average 3.78%–29.5% the system energy efficiency and reduce 15.7%–70.4% the performance penalty on different MPSoC scales.

With the fast development of inter/intra-chip optical interconnects, the gap between the data rates of electrical interconnects and optical interconnects is continuously increasing. Electrical–optical (E-O) interfaces and optical–electrical (O-E) interfaces are a pair of components that convert data between parallel electrical interconnects and serial optical interconnects. This paper holistically models and analyzes E-O and O-E interfaces in terms of energy consumption, area, and latency. Traditional interfaces, where data are converted between parallel and serial ports by serializers and deserializers (SerDes), are studied. A new type of E-O and O-E interface, which serializes and deserializes data by optical weaving technologies, are proposed alongside. Traditional interfaces will become a bottleneck for the further development of optical interconnects in the near future because of the high energy consumption and large area of SerDes necessitating new technologies. Our analysis shows that optical weaving interfaces have a better overall performance than traditional interfaces. For example, if there are 64 parallel electrical interconnects and four optical wavelengths, optical weaving interfaces can achieve a 81.6% improvement in energy consumption and a 40.8% improvement in area, compared with traditional interfaces.

With the fast development of processor chips, power-efficient, high-bandwidth, and low-latency interchip interconnects become more and more important. Studies show that the bandwidth of traditional parallel interconnects with low I/O clock frequencies will become bottlenecks in the near future. To solve this problem, two types of high-bandwidth interchip interconnects are developed. Low-swing differential electrical interconnects have widely been used in high-speed I/O designs. On the other hand, optical interconnects promise high bandwidth, low latency, and could improve the chip pin performance for manycore processors. They are becoming potential alternatives for electrical interconnects. This paper systematically models these two types of interconnects in terms of crosstalk noises, attenuation, and receiver sensitivities. Based on the proposed models, we developed optical and electrical interfaces and links (OEIL) and an analysis tool for OEIL. The OEIL can be used to analyze the energy consumption, bandwidth density, and latency of interconnects. Analytical models are verified by the results of published experiments. It shows that the optical interconnects have much higher bandwidth densities than the electrical interconnects. With this feature, the optical interconnects can significantly reduce I/O pin count compared with the electrical interconnects. For example, they can save at least 92% signal pins when connecting chips more than 25 cm (10 in) apart. The energy consumption of optical interconnects is comparable with that of electrical interconnects, and the latency of polymer waveguide-based optical interconnects is 18% less than that of electrical interconnect.

Chip I/O pins are an increasingly limited resource and significantly affect the performance, power and cost of multicore processors. Optical interconnects promise low power and high bandwidth, and are potential alternatives to electrical interconnects. This work systematically developed a set of analytical models for electrical and optical interconnects to study their structures, receiver sensitivities, crosstalk noises, and attenuations. We verified the models by published implementation results. The analytical models quantitatively identified the advantages of optical interconnects in terms of bandwidth, energy consumption, and transmission distance. We showed that optical interconnects can significantly reduce chip pin counts. For example, compared to electrical interconnects, optical interconnects can save at least 92% signal pins when connecting chips more than 25 cm (10 inches) apart.