Estimating power consumption is critical for hardware and software developers, and of the latter, particularly for OS programmers writing process schedulers. However, obtaining processor and system power consumption information can be non-trivial. Simulators are time consuming and prone to error. Power meters report whole-system consumption, but cannot give per-processor or per-thread information. More intrusive hardware instrumentation is possible, but such solutions are usually employed while designing the system, and are not meant for customer use.Given these difficulties, plus the current availability of some form of performance counters on virtually all platforms (even though such counters were initially designed for system bring-up, and not intended for general programmer consumption), we analytically derive functions for realtime estimation of processor and system power consumption using performance counter data on real hardware. Our model uses data gathered from microbenchmarks that capture potential application behavior. The model is independent of our test benchmarks, and thus we expect it to be well suited for future applications. We target chip multiprocessors, analyzing effects of shared resources and temperature on power estimation, leveraging our model to implement a simple, power-aware thread scheduler. The NAS and SPEC-OMP benchmarks shows a median error of 5.8% and 3.9%, respectively. SPEC 2006 shows a marginally higher median error of 7.2%.

Abstract-Performance, power, and temperature are now all first-order design constraints. Balancing power efficiency, thermal constraints, and performance requires some means to convey data about real-time power consumption and tem perature to intelligent resource managers. Resource managers can use this information to meet performance goals, maintain power budgets, and obey thermal constraints. Unfortunately, obtaining the required machine introspection is challenging.Most current chips provide no support for per-core power monitoring, and when support exists, it is not exposed to software. We present a methodology for deriving per-core power models using sampled performance counter values and temperature sensor readings. We develop application independent models for four different (four-to eight-core) platforms, validate their accuracy, and show how they can be used to guide scheduling decisions in power-aware resource managers. Model overhead is negligible, and estimations exhibit 1.1 %-5.2% per-suite median error on the NAS, SPEC OMP, and SPEC 2006 benchmarks (and 1.2%-4.4% overall). I. IN TRODUCTIONPower and temperature have joined performance as first order system design constraints. All three influence each other, and together they affect architectural and packaging choices. Power consumption characteristics further influence operating cost, reliability, battery lifetime, and device life time. Balancing power efficiency and thermal constraints with performance requires intelligent resource management, and achieving that balance requires real-time power con sumption and temperature information broken down accord ing to resource, together with software and hardware that can leverage such information to enforce management policies.One logical place to institute intelligent resource manage ment with respect to power, performance, and temperature for chip mUltiprocessor (CMP) systems is at the level of in dividual cores. Measuring run-time power of a single core is problematic, though. Current chips do not support it. Power meters only report total consumption for everything behind a single power cable, and even if such aggregate data were sufficient, the use of meters becomes completely infeasible as machines scale up: coordinating output and feedback from thousands of meters would require a separate (super) computing system. Cycle-level system simulators provide in depth information, but are extremely time consuming and 978-1-4244-7614-511 0/$26.00 ©20 10 IEEE prone to error. Power models implemented on top of the architectural abstractions in such simulators are inherently inaccurate [19], and are impossible to verify when attempt ing to assess new architectural designs. Hardware could be enhanced to measure the current and power draw of a CPU socket, but per-core measurement is difficult when cores share a power plane. Embedding measurement devices on chip is usually infeasible. Even when measurement facilities exist -e.g., the Intel Core i7 [16] features per-core power monitoring at the chip-level -they ar...

We develop real-time scheduling techniques for improving performance and energy for multiprogrammed workloads that scale nonuniformly with increasing thread counts. Multithreaded programs generally deliver higher throughput than single-threaded programs on chip multiprocessors, but performance gains from increasing threads decrease when there is contention for shared resources. We use analytic metrics to derive local search heuristics for creating efficient multiprogrammed, multithreaded workload schedules. Programs are allocated fewer cores than requested, and scheduled to space-share the CMP to improve global throughput. Our holistic approach attempts to co-schedule programs that complement each other with respect to shared resource consumption. We find application co-scheduling for performance and energy in a resource-aware manner achieves better results than solely targeting total throughput or concurrently co-scheduling all programs. Our schedulers improve overall energy delay (E*D) by a factor of 1.5 over time-multiplexed gang scheduling.

No abstract

No abstract

No abstract

No abstract

No abstract