Search for: All records

Smart Servers, Smarter Speed Scaling: A Decentralized Algorithm for Data Center Efficiency A team of researchers from Georgia Tech and the University of Minnesota has introduced a cutting-edge algorithm designed to optimize energy use in large-scale data centers. As detailed in their paper “Distributed Rate Scaling in Large-Scale Service Systems,” the team developed a decentralized method allowing each server to adjust its processing speed autonomously without the need for communication or knowledge of system-wide traffic. The algorithm uses idle time as a local signal to guide processing speed, ensuring that all servers converge toward a globally optimal performance rate. This innovation addresses a critical issue in modern computing infrastructure: balancing energy efficiency with performance under uncertainty and scale. The authors demonstrate that their approach not only stabilizes the system but achieves asymptotic optimality as the number of servers increases. The work is poised to significantly reduce energy consumption in data centers, which are projected to account for up to 8% of U.S. electricity use by 2030. 

Researchers have developed a novel model inspired by quantum switches to address the complexities of matching requests for entangled qubits in a discrete-time system. The study examines two types of arrivals: requests for entangled qubits between nodes and qubits supplied by nodes, which are subject to decoherence over time. Unlike classical queueing models, this system features server-less multiway matching and correlated abandonments, posing unique analytical challenges. By applying a max-weight policy, the researchers characterized the system’s stability using a two-time-scale fluid limit to account for qubit abandonments. They demonstrated that the max-weight policy is throughput optimal, outperforming nonidling policies under certain conditions. Intriguingly, the study revealed counterintuitive behavior: The longest request queue may grow temporarily, even in a stable system. These findings offer new insights into managing quantum-inspired systems with practical constraints, opening avenues for further research into quantum network optimization. 

We consider a load-balancing system composed of a fixed number of single-server queues operating under the well-known join-the-shortest queue policy and where jobs/customers are impatient and abandon if they do not receive service after some (random) amount of time. In this setting, we characterize the centered and appropriately scaled steady-state queue-length distribution (hereafter referred to as limiting distribution) in the limit as the abandonment rate goes to zero at the same time as the load either converges to one or is larger than one. Depending on the arrival, service, and abandonment rates, we observe three different regimes of operation that yield three different limiting distributions. The first regime is when the system is underloaded, and its load converges relatively slowly to one. In this case, abandonments do not affect the limiting distribution, and we obtain the same exponential distribution as in the system without abandonments. When the load converges to one faster, we have the second regime, where abandonments become significant. Here, the system undergoes a phase transition, and the limiting distribution is a truncated Gaussian. Further, the third regime is when the system is heavily overloaded, and so, the queue lengths are very large. In this case, we show that the limiting distribution converges to a normal distribution. To establish our results, we first prove a weaker form of state space collapse by providing a uniform bound on the second moment of the (unscaled) perpendicular component of the queue lengths, which shows that the system behaves like a single-server queue. We then use exponential Lyapunov functions to characterize the limiting distribution of the steady-state queue-length vector. Funding: This work was supported by the National Science Foundation [Grants CMMI-2140534 and EPCN-2144316]. 

We consider a large-scale parallel-server loss system with an unknown arrival rate, where each server is able to adjust its processing speed. The objective is to minimize the system cost, which consists of a power cost to maintain the servers' processing speeds and a quality of service cost depending on the tasks' processing times, among others. We draw on ideas from stochastic approximation to design a novel speed scaling algorithm and prove that the servers' processing speeds converge to the globally asymptotically optimum value. Curiously, the algorithm is fully distributed and does not require any communication between servers. Apart from the algorithm design, a key contribution of our approach lies in demonstrating how concepts from the stochastic approximation literature can be leveraged to effectively tackle learning problems in large-scale, distributed systems. En route, we also analyze the performance of a fully heterogeneous parallel-server loss system, where each server has a distinct processing speed, which might be of independent interest. 

We consider multi-armed bandit problems in social groups where in each individual has bounded memory and shares the common goal of learning the best arm/option. We say an individual learns the best option if eventually (ast→∞) it pulls only the arm with the highest expected reward. While this goal is provably impossible for an isolated individual due to bounded memory, we show that, in social groups, this goal can be achieved easily with the aid of social persuasion (i.e., communication) as long as the communication networks/graphs satisfy some mild conditions. In this work, we model and analyze a type of learning dynamics which are well-observed in social groups. Specifically, under the learning dynamics of interest, an individual sequentially decides on which arm to pull next based on not only its private reward feedback but also the suggestion provided by a randomly chosen neighbor. To deal with the interplay between the randomness in the rewards and in the social interaction, we employ the mean-field approximation method.Considering the possibility that the individuals in the networks may not be exchangeable when the communication networks are not cliques, we go beyond the classic mean-field techniques and apply a refined version of mean-field approximation. Notably, our results hold even if the communication graphs are highly sparse.