Luiz André Barroso

The Case for Energy-Proportional Computing

Energy-proportional designs would enable large energy savings in servers, potentially doubling their efficiency in real-life use. Achieving energy proportionality will require significant improvements in the energy usage profile of every system component, particularly the memory and disk subsystems.

Power provisioning for a warehouse-sized computer

Large-scale Internet services require a computing infrastructure that can beappropriately described as a warehouse-sized computing system. The cost ofbuilding datacenter facilities capable of delivering a given power capacity tosuch a computer can rival the recurring energy consumption costs themselves.Therefore, there are strong economic incentives to operate facilities as closeas possible to maximum capacity, so that the non-recurring facility costs canbe best amortized. That is difficult to achieve in practice because ofuncertainties in equipment power ratings and because power consumption tends tovary significantly with the actual computing activity. Effective powerprovisioning strategies are needed to determine how much computing equipmentcan be safely and efficiently hosted within a given power budget. In this paper we present the aggregate power usage characteristics of largecollections of servers (up to 15 thousand) for different classes ofapplications over a period of approximately six months. Those observationsallow us to evaluate opportunities for maximizing the use of the deployed powercapacity of datacenters, and assess the risks of over-subscribing it. We findthat even in well-tuned applications there is a noticeable gap (7 - 16%)between achieved and theoretical aggregate peak power usage at the clusterlevel (thousands of servers). The gap grows to almost 40% in wholedatacenters. This headroom can be used to deploy additional compute equipmentwithin the same power budget with minimal risk of exceeding it. We use ourmodeling framework to estimate the potential of power management schemes toreduce peak power and energy usage. We find that the opportunities for powerand energy savings are significant, but greater at the cluster-level (thousandsof servers) than at the rack-level (tens). Finally we argue that systems needto be power efficient across the activity range, and not only at peakperformance levels.

Web search for a planet: The Google cluster architecture

Amenable to extensive parallelization, Googles web search application lets different queries run on different processors and, by partitioning the overall index, also lets a single query use multiple processors. to handle this workload, Googless architecture features clusters of more than 15,000 commodity-class PCs with fault tolerant software. This architecture achieves superior performance at a fraction of the cost of a system built from fewer, but more expensive, high-end servers.Amenable to extensive parallelization, Googles web searchapplication lets different queries run on different processors and,by partitioning the overall index, also lets a single query usemultiple ...

Piranha: a scalable architecture based on single-chip multiprocessing

This paper describes the Piranha system, a research prototype being developed at Compaq that aggressively exploits chip multiprocessing by integrating eight simple Alpha processor cores along with a two-level cache hierarchy onto a single chip. Piranha also integrates further on-chip functionality to allow for scalable multiprocessor configurations to be built in a glueless and modular fashion. The use of simple processor cores combined with an industry-standard ASIC design methodology allow us to complete our prototype within a short time-frame, with a team size and investment that are an order of magnitude smaller than that of a commercial microprocessor. Our detailed simulation results show that while each Piranha processor core is substantially slower than an aggressive next-generation processor, the integration of eight cores onto a single chip allows Piranha to outperform next-generation processors by up to 2.9 times (on a per chip basis) on important workloads such as OLTP. This performance advantage can approach a factor of five by using full-custom instead of ASIC logic. In addition to exploiting chip multiprocessing, the Piranha prototype incorporates several other unique design choices including a shared second-level cache with no inclusion, a highly optimized cache coherence protocol, and a novel I/O architecture.

Memory system characterization of commercial workloads

Commercial applications such as databases and Web servers constitute the largest and fastest-growing segment of the market for multiprocessor servers. Ongoing innovations in disk subsystems, along with the ever increasing gap between processor and memory speeds, have elevated memory system design as the critical performance factor for such workloads. However, most current server designs have been optimized to perform well on scientific and engineering workloads, potentially leading to design decisions that are non-ideal for commercial applications. The above problem is exacerbated by the lack of information on the performance requirements of commercial workloads, the lack of available applications for widespread study, and the fact that most representative applications are too large and complex to serve as suitable benchmarks for evaluating trade-offs in the design of processors and servers.This paper presents a detailed performance study of three important classes of commercial workloads: online transaction processing (OLTP), decision support systems (DSS), and Web index search. We use the Oracle commercial database engine for our OLTP and DSS workloads, and the AltaVista search engine for our Web index search workload. This study characterizes the memory system behavior of these workloads through a large number of architectural experiments on Alpha multiprocessors augmented with full system simulations to determine the impact of architectural trends. We also identify a set of simplifications that make these workloads more amenable to monitoring and simulation without affecting representative memory system behavior. We observe that systems optimized for OLTP versus DSS and index search workloads may lead to diverging designs, specifically in the size and speed requirements for off-chip caches.

Power management of online data-intensive services

Much of the success of the Internet services model can be attributed to the popularity of a class of workloads that we call Online Data-Intensive (OLDI) services. These work-loads perform significant computing over massive data sets per user request but, unlike their offline counterparts (such as MapReduce computations), they require responsiveness in the sub-second time scale at high request rates. Large search products, online advertising, and machine translation are examples of workloads in this class. Although the load in OLDI services can vary widely during the day, their energy consumption sees little variance due to the lack of energy proportionality of the underlying machinery. The scale and latency sensitivity of OLDI workloads also make them a challenging target for power management techniques. We investigate what, if anything, can be done to make OLDI systems more energy-proportional. Specifically, we evaluate the applicability of active and idle low-power modes to reduce the power consumed by the primary server components (processor, memory, and disk), while maintaining tight response time constraints, particularly on 95th-percentile latency. Using Web search as a representative example of this workload class, we first characterize a production Web search workload at cluster-wide scale. We provide a fine-grain characterization and expose the opportunity for power savings using low-power modes of each primary server component. Second, we develop and validate a performance model to evaluate the impact of processor- and memory-based low-power modes on the search latency distribution and consider the benefit of current and foreseeable low-power modes. Our results highlight the challenges of power management for this class of workloads. In contrast to other server workloads, for which idle low-power modes have shown great promise, for OLDI workloads we find that energy-proportionality with acceptable query latency can only be achieved using coordinated, full-system active low-power modes.

The Price of Performance

In the late 1990s, our research group at DEC was one of a growing number of teams advocating the CMP (chip multiprocessor) as an alternative to highly complex single-threaded CPUs. We were designing the Piranha system,1 which was a radical point in the CMP design space in that we used very simple cores (similar to the early RISC designs of the late ’80s) to provide a higher level of thread-level parallelism. Our main goal was to achieve the best commercial workload performance for a given silicon budget. Today, in developing Google’s computing infrastructure, our focus is broader than performance alone. The merits of a particular architecture are measured by answering the following question: Are you able to afford the computational capacity you need? The high-computational demands that are inherent in most of Google’s services have led us to develop a deep understanding of the overall cost of computing, and continually to look for hardware/software designs that optimize performance per unit of cost.

Performance of database workloads on shared-memory systems with out-of-order processors

Database applications such as online transaction processing (OLTP) and decision support systems (DSS) constitute the largest and fastest-growing segment of the market for multiprocessor servers. However, most current system designs have been optimized to perform well on scientific and engineering workloads. Given the radically different behavior of database workloads (especially OLTP), it is important to re-evaluate key system design decisions in the context of this important class of applications.This paper examines the behavior of database workloads on shared-memory multiprocessors with aggressive out-of-order processors, and considers simple optimizations that can provide further performance improvements. Our study is based on detailed simulations of the Oracle commercial database engine. The results show that the combination of out-of-order execution and multiple instruction issue is indeed effective in improving performance of database workloads, providing gains of 1.5 and 2.6 times over an in-order single-issue processor for OLTP and DSS, respectively. In addition, speculative techniques enable optimized implementations of memory consistency models that significantly improve the performance of stricter consistency models, bringing the performance to within 10--15% of the performance of more relaxed models.The second part of our study focuses on the more challenging OLTP workload. We show that an instruction stream buffer is effective in reducing the remaining instruction stalls in OLTP, providing a 17% reduction in execution time (approaching a perfect instruction cache to within 15%). Furthermore, our characterization shows that a large fraction of the data communication misses in OLTP exhibit migratory behavior; our preliminary results show that software prefetch and writeback/flush hints can be used for this data to further reduce execution time by 12%.

The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines, Second edition

Abstract As computation continues to move into the cloud, the computing platform of interest no longer resembles a pizza box or a refrigerator, but a warehouse full of computers. These new large datacenters are quite different from traditional hosting facilities of earlier times and cannot be viewed simply as a collection of co-located servers. Large portions of the hardware and software resources in these facilities must work in concert to efficiently deliver good levels of Internet service performance, something that can only be achieved by a holistic approach to their design and deployment. In other words, we must treat the datacenter itself as one massive warehouse-scale computer (WSC). We describe the architecture of WSCs, the main factors influencing their design, operation, and cost structure, and the characteristics of their software base. We hope it will be useful to architects and programmers of today’s WSCs, as well as those of future many-core platforms which may one day implement the equivale...

Towards energy proportionality for large-scale latency-critical workloads

Reducing the energy footprint of warehouse-scale computer (WSC) systems is key to their affordability, yet difficult to achieve in practice. The lack of energy proportionality of typical WSC hardware and the fact that important workloads (such as search) require all servers to remain up regardless of traffic intensity renders existing power management techniques ineffective at reducing WSC energy use. We present PEGASUS, a feedback-based controller that significantly improves the energy proportionality of WSC systems, as demonstrated by a real implementation in a Google search cluster. PEGASUS uses request latency statistics to dynamically adjust server power management limits in a fine-grain manner, running each server just fast enough to meet global service-level latency objectives. In large cluster experiments, PEGASUS reduces power consumption by up to 20%. We also estimate that a distributed version of PEGASUS can nearly double these savings.