A large-scale online discrete optimization model is formulates and a new online sampling optimization (OSO) algorithm is developed that anticipates future demand by repeatedly simulating future arrivals and reoptimizing decisions over time, demonstrating the real-world impact of optimization in cloud infrastructure management.
Abstract
Online Rack Placement in Large-Scale Data Centers: Online Sampling Optimization and Deployment
Data centers have grown into major components of global supply chains. This paper develops, deploys, and assesses an optimization algorithm to improve how large-scale data centers place incoming server racks dynamically while balancing space, power, cooling, and reliability constraints. Poor placement decisions can leave valuable resources stranded, leading to high costs and reduced operating resilience. This paper formulates a large-scale online discrete optimization model and develops a new online sampling optimization (OSO) algorithm that anticipates future demand by repeatedly simulating future arrivals and reoptimizing decisions over time. Theoretical results provide performance guarantees, and computational results show its benefits against state-of-the-art reoptimization methods. The system was implemented as a decision-support tool and deployed across Microsoft’s global fleet of data centers. Using postdeployment data, the paper shows that adoption of the tool reduced power stranding by one to three percentage points. At Microsoft’s scale, these improvements translate into substantial financial savings and meaningful reductions in greenhouse gas emissions, demonstrating the real-world impact of optimization in cloud infrastructure management.
Virtual Network Functions (VNFs) deployed on commodity servers exhibit failure rates two orders of magnitude higher than carrier-grade hardware, rendering service function chain (SFC) reliability a critical concern. Existing solutions are confined to offline static scenarios and single-instance serial models, which create single-point bottlenecks and constrain network service provider (NSP) revenue. This paper proposes GRM-SFCOD, which is a General Reliability-aware Multi-replica SFC Online Deployment algorithm. We introduce a primary-backup replica pool that parallelizes VNFs into lightweight instances while provisioning redundant backups for fault tolerance. The online deployment problem is formulated as a Mixed-Integer Nonlinear Program (MINLP) that maximizes NSP revenue; its NP-hardness is proved via reduction from the Multidimensional Knapsack Problem. To enable efficient online decisions, we design a dynamic pricing mechanism with exponentially scaling resource prices, decomposing the problem into (i) replica allocation based on marginal reliability gain and (ii) VNF placement via an improved Genetic Algorithm. Simulations demonstrate a ∼9.68% lower deployment cost compared to uniform-allocation baselines, ∼17% higher resource utilization than Best-Fit heuristics, and an attainment of ∼81% of offline optimal revenue with orders-of-magnitude lower computational overhead.
Haitong Gu, Bin Guo, Jun Dong et al.· Electronics· 0 citations
Efficient transient scheduling of cluster tools is critical to stable fab-level operations in high-mix, low-volume semiconductor manufacturing. This study investigates the transient scheduling problem of dual-armed cluster tools (DACTs) processing two wafer types concurrently. From an engineering-management perspective, the problem is important because wafer release and robot sequencing decisions affect tool utilization, completion-time predictability, and fab-level responsiveness. A mixed-integer programming model is developed to jointly optimize wafer release order and robot task sequence with the objective of minimizing total completion time. To address the computational intractability of medium- and large-scale instances, a two-combined swap sequence (2-CSS) is proposed. Numerical results show that 2-CSS substantially reduces computation time while incurring only a 5.31% average increase in total completion time. The results also show that DACTs provide substantial throughput benefits over single-armed configurations, with improvements of up to 9.51%. Robustness analysis further shows that the observed dual-arm advantage persists across varying robot loading/unloading times and wafer residency time constraint-to-processing time ratios, while 2-CSS maintains small optimality gaps as the loading/unloading time and the processing time of a selected step vary, even when the system bottleneck shifts between processing steps. These findings can help fab managers evaluate transient schedules more quickly and coordinate tool operations more effectively in high-mix production.
Zhenyan Wu, Fajun Yang, Chao Li et al.· IEEE transactions on enginee...· 0 citations
JointCloud environments, including multi-cloud and federated cloud systems, increasingly rely on high-performance networks (HPNs) to support large-scale cross-cloud data transfers. In such settings, advance bandwidth reservation with timedependent pricing is essential for cost-efficient and predictable data movement, where the transfer cost depends on both dynamic link prices and path reconfiguration overhead. This paper investigates the optimal scheduling of the VPFB BRR-MinC, where VPFB (Variable Path, Fixed Bandwidth) allows routing paths to change across time slots while maintaining a constant reserved bandwidth, and BRR-MinC seeks a minimum-cost schedule for deadline-constrained data transfers. We formalize a timedependent cost model incorporating slot-varying edge weights and switching penalties, and prove that the problem is NPcomplete. To address the temporal coupling introduced by switching costs, we develop a segmentation-based dynamic programming framework and propose a scalable heuristic, Heu-VPFB-MinC-TD-S. Simulation results on an ESnet-inspired topology show that the proposed method achieves identical feasibility while reducing total transfer cost compared with a greedy baseline, at the expense of moderate additional runtime. These results demonstrate the effectiveness of segmentation-aware optimization for cost-efficient cross-cloud data transfer in JointCloud systems.
Liudong Zuo, Pan Lai, Michelle Zhu et al.· Fall Joint Computer Conferen...· 0 citations
The rapid growth of LLM inference is shifting sustainability concerns from one-time training to continuous serving, where infrastructure decisions shape energy use, carbon emissions, water consumption, and service quality. Yet operators often need to compare deployment alternatives before large-scale infrastructure is built, making direct measurement costly, slow, and sometimes infeasible. We present InFactPlanner, a trace-driven decision-support framework for what-if analysis of sustainable AI data center deployment for LLM inference across single and geo-distributed sites. InFactPlanner combines query traces, hardware-model profiles, candidate site configurations, PUE/WUE parameters, renewable generation models, and time-varying grid carbon intensity to estimate power, energy, carbon emissions, water use, latency, and server utilization. The framework abstracts low-level serving effects into configurable hardware-model profiles, enabling rapid comparison of site selection, capacity placement, hardware, model, renewable integration, and routing choices. We validate the energy accounting pipeline by reproducing reference LLM inference energy estimates with less than 10% deviation, evaluate scalability across multiple data centers and server counts, and demonstrate scenario-driven decision analyses for hardware selection, renewable placement, geographic deployment, and carbon-aware routing. Our results show that sustainability-optimal choices can differ from latency-optimal ones, and that the carbon value of deployment depends strongly on the local grid mix.
Nicoletta Tsiopani, Moysis Symeonides, G. Pallis et al.· 0 citations
We study online service with one maximum-waiting-time charge per service batch. The persistent server endpoint prevents a phase-by-phase comparison with the offline optimum: an offline schedule may merge many online phases, share movement globally, and finish at unrelated endpoints. Our main contribution is a metric-independent \emph{group--trajectory certificate framework} that restores such a comparison. For ordered request groups in disjoint time windows, a certificate value is bounded both by the window length and by the metric Steiner cost of the group. After normalizing the offline schedule into consecutive arrival blocks, strictly interior groups are charged to offline delay, while boundary groups induce connectors of congestion at most two along the offline trajectory. One color class therefore has certificate sum at most $2\OPT$; a parity decomposition yields $\sum_h C_h\le4\OPT$. Consequently, any phase rule whose cost is at most $\alpha C_h$ is $4\alpha$-competitive. For visible service, this theorem yields deterministic ratios $10$ on a line, $12$ on a weighted tree, and $20$ on an arbitrary finite metric; the last algorithm is polynomial and uses a phase-local terminal-MST envelope, while an exact metric-Steiner oracle gives ratio $12$. Structurally, elective and automatic schedules can have different event structures but equal offline optimal values. The common value is computable exactly in polynomial time on lines and explicitly represented weighted trees, whereas exact optimization on arbitrary finite metrics is NP-hard. Finally, we use spatial blindness---announced requests whose locations are revealed only when visited---as a stress test: dyadic exploration preserves a constant ratio on a known finite line, while a single hidden request on a star forces a loss linear in its degree.
Tianhan Lu, Runtian Ren, Shengcai Liu et al.· 0 citations
Order picking is one of the most costly activities in manual picker-to-parts warehouses, and routing quality has a direct impact on travel effort and operational efficiency. However, commercial warehouse management systems (WMSs) still often rely on simple rule-based policies because exact optimization is difficult to reconcile with real-time execution requirements. This paper presents a WMS-compatible graph-based routing method for manual warehouses based on mission-dependent graph reduction, shortest-path computation, and sequencing optimization over mission-relevant locations. Starting from the physical warehouse graph, the proposed method builds a compact reduced representation that preserves shortest-path distances while significantly decreasing the online computational burden, enabling seamless integration into existing WMSs and real-time operation. The method is validated on a real household-goods distribution center and compared with both practical rule-based routing policies typically adopted in commercial WMSs and exact optimization benchmarks. Results on both synthetic missions and real warehouse orders show substantial travel-distance reductions with runtimes fully compatible with online warehouse operation.
Federico Signorile, Raffaele Carli, M. Gorgoglione et al.· International Conference on...· 0 citations