Stefan Larson, Anish Mahendran, Joseph J. Peper, Christopher Clarke, Andrew Lee, Parker Hill, Jonathan K. Kummerfeld, Kevin Leach, Michael A. Laurenzano, Lingjia Tang, Jason Mars. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

We introduce Caliper , a technique for accurately estimating performance interference occurring in shared servers. Caliper overcomes the limitations of prior approaches by leveraging a micro-experiment-based technique. In contrast to state-of-the-art approaches that focus on periodically pausing co-running applications to estimate slowdown, Caliper utilizes a strategic phase-triggered technique to capture interference due to co-location. This enables Caliper to orchestrate an accurate and low-overhead interference estimation technique that can be readily deployed in existing production systems. We evaluate Caliper for a broad spectrum of workload scenarios, demonstrating its ability to seamlessly support up to 16 applications running simultaneously and outperform the state-of-the-art approaches.

A key focus of recent work in our community has been on devising increasingly sophisticated acceleration devices for deep neural network (DNN) computation, especially for networks driven by convolution layers. Yet, despite the promise of substantial improvements in performance and energy consumption offered by these approaches, general purpose computing is not going away because its traditional well-understood programming model and continued wide deployment. Therefore, the question arises as to what can be done, if anything, to evolve conventional CPUs to accommodate efficient deep neural network computation.

Cloud-scale datacenter management systems utilize virtualization to provide performance isolation while maximizing the utilization of the underlying hardware infrastructure. However, virtualization does not provide complete performance isolation as Virtual Machines (VMs) still compete for nonreservable shared resources (like caches, network, I/O bandwidth etc.) This becomes highly challenging to address in datacenter environments housing tens of thousands of VMs, causing degradation in application performance. Addressing this problem for production datacenters requires a non-intrusive scalable solution that 1) detects performance intrusion and 2) investigates both the intrusive VMs causing interference, as well as the resource(s) for which the VMs are competing for. To address this problem, this paper introduces Proctor, a real time, lightweight and scalable analytics fabric that detects performance intrusive VMs and identifies its root causes from among the arbitrary VMs running in shared datacenters across 4 key hardware resources - network, I/O, cache, and CPU. Proctor is based on a robust statistical approach that requires no special profiling phases, standing in stark contrast to a wide body of prior work that assumes pre-acquisition of application level information prior to its execution. By detecting performance degradation and identifying the root cause VMs and their metrics, Proctor can be utilized to dramatically improve the performance outcomes of applications executing in large-scale datacenters. From our experiments, we are able to show that when we deploy Proctor in a datacenter housing a mix of I/O, network, compute and cache-sensitive applications, it is able to effectively pinpoint performance intrusive VMs. Further, we observe that when Proctor is applied with migration, the application-level Quality-of-Service improves by an average of 2.2× as compared to systems which are unable to detect, identify and pinpoint performance intrusion and their root causes.

Deep neural networks (DNNs) are key computational building blocks for emerging classes of web services that interact in real time with users via voice, images and video inputs. Although GPUs have gained popularity as a key accelerator platform for deep learning workloads, the increasing demand for DNN computation leaves a significant gap between the compute capabilities of GPU-enabled datacenters and the compute needed to service demand.

Reducing the long tail of the query latency distribution in modern warehouse scale computers is critical for improving performance and quality of service (QoS) of workloads such as Web Search and Memcached. Traditional turbo boost increases a processor’s voltage and frequency during a coarse-grained sliding window, boosting all queries that are processed during that window. However, the inability of such a technique to pinpoint tail queries for boosting limits its tail reduction benefit. In this work, we propose Adrenaline , an approach to leverage finer-granularity (tens of nanoseconds) voltage boosting to effectively rein in the tail latency with query-level precision. Two key insights underlie this work. First, emerging finer granularity voltage/frequency boosting is an enabling mechanism for intelligent allocation of the power budget to precisely boost only the queries that contribute to the tail latency; second, per-query characteristics can be used to design indicators for proactively pinpointing these queries, triggering boosting accordingly. Based on these insights, Adrenaline effectively pinpoints and boosts queries that are likely to increase the tail distribution and can reap more benefit from the voltage/frequency boost. By evaluating under various workload configurations, we demonstrate the effectiveness of our methodology. We achieve up to a 2.50 × tail latency improvement for Memcached and up to a 3.03 × for Web Search over coarse-grained dynamic voltage and frequency scaling (DVFS) given a fixed boosting power budget. When optimizing for energy reduction, Adrenaline achieves up to a 1.81 × improvement for Memcached and up to a 1.99 × for Web Search over coarse-grained DVFS. By using the carefully chosen boost thresholds, Adrenaline further improves the tail latency reduction to 4.82 × over coarse-grained DVFS.

Being able to identify low-performing students early in the term may help instructors intervene or differently allocate course resources. Prior work in CS1 has demonstrated that clicker correctness in Peer Instruction courses correlates with exam outcomes and, separately, that machine learning models can be built based on early-term programming assessments. This work aims to combine the best elements of each of these approaches. We offer a methodology for creating models, based on in-class clicker questions, to predict cross-term student performance. In as early as week 3 in a 12-week CS1 course, this model is capable of correctly predicting students as being in danger of failing, or not, for 70% of the students, with only 17% of students misclassified as not at-risk when at-risk. Additional measures to ensure more broad applicability of the methodology, along with possible limitations, are explored.

On-core microarchitectural structures consume significant portions of a processor's power budget. However, depending on application characteristics, those structures do not always provide (much) performance benefit. While timeout-based power gating techniques have been leveraged for underutilized cores and inactive functional units, these techniques have not directly translated to high-activity units such as vector processing units, complex branch predictors, and caches. The performance benefit provided by these units does not necessarily correspond with unit activity, but instead is a function of application characteristics. This work introduces PowerChop, a novel technique that leverages the unique capabilities of HW/SW co-designed hybrid processors to enact unit-level power management at the application phase level. PowerChop adds two small additional hardware units to facilitate phase identification and triggering different power states, enabling the software layer to cheaply track, predict and take advantage of varying unit criticality across application phases by powering gating units that are not needed for performant execution. Through detailed experimentation, we find that PowerChop significantly decreases power consumption, reducing the leakage power of a hybrid server processor by 9% on average (up to 33%) and a hybrid mobile processor by 19% (up to 40%) while introducing just 2% slowdown.

This paper presents the first comprehensive study of the performance, power and energy consumption of the Applied-Micro X-Gene, the first commercially available 64-bit ARMv8 platform, for HPC workloads. Our study includes a detailed comparison of the X-Gene to three other architectural design points common in HPC systems. Across these platforms, we perform careful measurements across 400+ workloads, covering different application domains, parallelization models, floating-point precision models and memory intensities. We find that the X-Gene has an average of 1.2× better energy consumption than an Intel Sandy Bridge, a design commonly found in HPC installations, while the Sandy Bridge is an average of 2.3× faster than X-Gene. Precisely quantifying the causes of performance and energy differences between two platforms is an important but challenging problem that is often addressed via detailed simulation, an approach that has limited ability to scale up to full applications and broad workload mixes. Instead, this paper adopts a statistical framework called Partial Least Squares (PLS) Path Modeling to solve this problem. PLS Path Modeling allows us to capture complex cause-effect relationships and difficult-to-measure performance concepts relating to the effectiveness of architectural units and subsystems in improving application performance using readily available hardware counter measurements. We use PLS Path Modeling to quantify the causes of the performance differences between X-Gene and Sandy Bridge in the HPC domain, finding that the performance of the memory subsystem is the dominant cause of these differences.

Cache capacity and memory bandwidth play critical roles in application performance, particularly for data-intensive applications from domains that include machine learning, numerical analysis, and data mining. Many of these applications are also tolerant to imprecise inputs and have loose constraints on the quality of output, making them ideal candidates for approximate computing. This paper introduces a novel approximate computing technique that decouples the format of data in the memory hierarchy from the format of data in the compute subsystem to significantly reduce the cost of storing and moving bits throughout the memory hierarchy and improve application performance. This asymmetric compute-memory extension to conventional architectures, ACME, adds two new instruction classes to the ISA - load-concise and store-concise - along with three small functional units to the micro-architecture to support these instructions. ACME does not affect exact execution of applications and comes into play only when concise memory operations are used. Through detailed experimentation we find that ACME is very effective at trading result accuracy for improved application performance. Our results show that ACME achieves a 1.3x speedup (up to 1.8x) while maintaining 99% accuracy, or a 1.1x speedup while maintaining 99.999% accuracy. Moreover, our approach incurs negligible area and power overheads, adding just 0.005% area and 0.1% power to a conventional modern architecture.

The class of optimizations characterized by manipulating a loop's interaction space for improved cache locality and reuse (i.e, cache tiling/blocking/strip mine and interchange) are static optimizations requiring a priori information about the microarchitectural and runtime environment of an application binary. However, particularly in datacenter environments, deployed applications face numerous dynamic environments over their lifetimes. As a result, this class of optimizations can result in sub-optimal performance due to the inability to flexibly adapt iteration spaces as cache conditions change at runtime. This paper introduces continuous shape shifiting, a compilation approach that removes the risks of cache tiling optimizations by dynamically rewriting (and reshaping) deployed, running application code. To realize continuous shape shifting, we present ShapeShifter, a framework for continuous monitoring of co-running applications and their runtime environments to reshape loop iteration spaces and pinpoint near-optimal loop tile configurations. Upon identifying a need for reshaping, a new tiling approach is quickly constructed for the application, new code is dynamically generated and is then seamlessly stitched into the running application with near-zero overhead. Our evaluation on a wide spectrum of runtime scenarios demonstrates that ShapeShifter achieves an average of 10-40% performance improvement (up to 2.4 χ) on real systems depending on the runtime environment compared to an oracle static loop tiling baseline.

Understanding dynamic program behavior is critical in many stages of the software development lifecycle, for purposes as diverse as optimization, debugging, testing, and security. This paper focuses on the problem of predicting dynamic program behavior statically. We introduce a novel technique to statically identify hot paths that leverages emerging deep learning techniques to take advantage of their ability to learn subtle, complex relationships between sequences of inputs. This approach maps well to the problem of identifying the behavior of sequences of basic blocks in program execution. Our technique is also designed to operate on the compiler's intermediate representation (IR), as opposed to the approaches taken by prior techniques that have focused primarily on source code, giving our approach language-independence. We describe the pitfalls of conventional metrics used for hot path prediction such as accuracy, and motivate the use of Area Under the Receiver Operating Characteristic curve (AUROC). Through a thorough evaluation of our technique on complex applications that include the SPEC CPU2006 benchmarks, we show that our approach achieves an AUROC of 0.85.

This paper introduces Input Responsive Approximation (IRA), an approach that uses a canary input — a small program input carefully constructed to capture the intrinsic properties of the original input — to automatically control how program approximation is applied on an input-by-input basis. Motivating this approach is the observation that many of the prior techniques focusing on choosing how to approximate arrive at conservative decisions by discounting substantial differences between inputs when applying approximation. The main challenges in overcoming this limitation lie in making the choice of how to approximate both effectively (e.g., the fastest approximation that meets a particular accuracy target) and rapidly for every input. With IRA, each time the approximate program is run, a canary input is constructed and used dynamically to quickly test a spectrum of approximation alternatives. Based on these runtime tests, the approximation that best fits the desired accuracy constraints is selected and applied to the full input to produce an approximate result. We use IRA to select and parameterize mixes of four approximation techniques from the literature for a range of 13 image processing, machine learning, and data mining applications. Our results demonstrate that IRA significantly outperforms prior approaches, delivering an average of 10.2× speedup over exact execution while minimizing accuracy losses in program outputs.

Summary The current state of practice in supercomputer resource allocation places jobs from different users on disjoint nodes both in terms of time and space. While this approach largely guarantees that jobs from different users do not degrade one another's performance, it does so at high cost to system throughput and energy efficiency. This focused study presents job striping, a technique that significantly increases performance over the current allocation mechanism by colocating pairs of jobs from different users on a shared set of nodes. To evaluate the potential of job striping in large‐scale environments, the experiments are run at the scale of 128 nodes on the state‐of‐the‐art Gordon supercomputer. Across all pairings of 1024 process network‐attached storage parallel benchmarks, job striping increases mean throughput by 26% and mean energy efficiency by 22%. On pairings of the real applications Gyrokinetic Toroidal Code (GTC), Large‐scale Atomic/Molecular Massively Parallel Simulator (LAMMPS), and MIMD Lattice Computation (MILC) at equal scale, job striping improves average throughput by 12% and mean energy efficiency by 11%. In addition, the study provides a simple set of heuristics for avoiding low performing application pairs. Copyright © 2013 John Wiley & Sons, Ltd.

Summary This article presents Green Queue, a production quality tracing and analysis framework for implementing application aware dynamic voltage and frequency scaling (DVFS) for message passing interface applications in high performance computing. Green Queue makes use of both intertask and intratask DVFS techniques. The intertask technique targets applications where the workload is imbalanced by reducing CPU clock frequency and therefore power draw for ranks with lighter workloads. The intratask technique targets balanced workloads where all tasks are synchronously running the same code. The strategy identifies program phases and selects the energy‐optimal frequency for each by predicting power and measuring the performance responses of each phase to frequency changes. The success of these techniques is evaluated on 1024 cores on Gordon, a supercomputer at the San Diego Supercomputer Center built using Intel Xeon E5‐2670 (Sandybridge) processors. Green Queue achieves up to 21% and 32% energy savings for the intratask and intertask DVFS strategies, respectively. Copyright © 2013 John Wiley & Sons, Ltd.

As user demand scales for intelligent personal assistants (IPAs) such as Apple’s Siri, Google’s Google Now, and Microsoft’s Cortana, we are approaching the computational limits of current datacenter (DC) architectures. It is an open question how future server architectures should evolve to enable this emerging class of applications, and the lack of an open-source IPA workload is an obstacle in addressing this question. In this article, we present the design of Sirius, an open end-to-end IPA Web-service application that accepts queries in the form of voice and images, and responds with natural language. We then use this workload to investigate the implications of four points in the design space of future accelerator-based server architectures spanning traditional CPUs, GPUs, manycore throughput co-processors, and FPGAs. To investigate future server designs for Sirius, we decompose Sirius into a suite of eight benchmarks (Sirius Suite) comprising the computationally intensive bottlenecks of Sirius. We port Sirius Suite to a spectrum of accelerator platforms and use the performance and power trade-offs across these platforms to perform a total cost of ownership (TCO) analysis of various server design points. In our study, we find that accelerators are critical for the future scalability of IPA services. Our results show that GPU- and FPGA-accelerated servers improve the query latency on average by 8.5× and 15×, respectively. For a given throughput, GPU- and FPGA-accelerated servers can reduce the TCO of DCs by 2.3× and 1.3×, respectively.

Modern processors widely use hardware prefetching to hide memory latency. While aggressive hardware prefetchers can improve performance significantly for some applications, they can limit the overall performance in highly-utilized multicore processors by saturating the offchip bandwidth and wasting last-level cache capacity. Co-executing applications can slowdown due to contention over these shared resources. This work introduces Adaptive Resource Efficient Prefetching (AREP) -- a runtime framework that dynamically combines software prefetching and hardware prefetching to maximize throughput in highly utilized multicore processors. AREP achieves better performance by prefetching data in a resource efficient way -- conserving offchip-bandwidth and last-level cache capacity with accurate prefetching and by applying cache-bypassing when possible. AREP dynamically explores a mix of hardware/software prefetching policies, then selects and applies the best performing policy. AREP is phase-aware and re-explores (at runtime) for the best prefetching policy at phase boundaries. A multitude of experiments with workload mixes and parallel applications on a modern high performance multicore show that AREP can increase throughput by up to 49% (8.1% on average). This is complemented by improved fairness, resulting in average quality of service above 94%.

As user demand scales for intelligent personal assistants (IPAs) such as Apple's Siri, Google's Google Now, and Microsoft's Cortana, we are approaching the computational limits of current datacenter architectures. It is an open question how future server architectures should evolve to enable this emerging class of applications, and the lack of an open-source IPA workload is an obstacle in addressing this question. In this paper, we present the design of Sirius, an open end-to-end IPA web-service application that accepts queries in the form of voice and images, and responds with natural language. We then use this workload to investigate the implications of four points in the design space of future accelerator-based server architectures spanning traditional CPUs, GPUs, manycore throughput co-processors, and FPGAs.

Reducing the long tail of the query latency distribution in modern warehouse scale computers is critical for improving performance and quality of service of workloads such as Web Search and Memcached. Traditional turbo boost increases a processor's voltage and frequency during a coarse-grain sliding window, boosting all queries that are processed during that window. However, the inability of such a technique to pinpoint tail queries for boosting limits its tail reduction benefit. In this work, we propose Adrenaline, an approach to leverage finer granularity, 10's of nanoseconds, voltage boosting to effectively rein in the tail latency with query-level precision. Two key insights underlie this work. First, emerging finer granularity voltage/frequency boosting is an enabling mechanism for intelligent allocation of the power budget to precisely boost only the queries that contribute to the tail latency; and second, per-query characteristics can be used to design indicators for proactively pinpointing these queries, triggering boosting accordingly. Based on these insights, Adrenaline effectively pinpoints and boosts queries that are likely to increase the tail distribution and can reap more benefit from the voltage/frequency boost. By evaluating under various workload configurations, we demonstrate the effectiveness of our methodology. We achieve up to a 2.50x tail latency improvement for Memcached and up to a 3.03x for Web Search over coarse-grained DVFS given a fixed boosting power budget. When optimizing for energy reduction, Adrenaline achieves up to a 1.81x improvement for Memcached and up to a 1.99x for Web Search over coarse-grained DVFS.

Heterogeneous multicore architectures have the potential to improve energy efficiency by integrating power-efficient wimpy cores with high-performing brawny cores. However, it is an open question as how to deliver energy reduction while ensuring the quality of service (QoS) of latency-sensitive web-services running on such heterogeneous multicores in warehouse-scale computers (WSCs). In this work, we first investigate the implications of heterogeneous multicores in WSCs and show that directly adopting heterogeneous multicores without re-designing the software stack to provide QoS management leads to significant QoS violations. We then present Octopus-Man, a novel QoS-aware task management solution that dynamically maps latency-sensitive tasks to the least power-hungry processing resources that are sufficient to meet the QoS requirements. Using carefully-designed feedback-control mechanisms, Octopus-Man addresses critical challenges that emerge due to uncertainties in workload fluctuations and adaptation dynamics in a real system. Our evaluation using web-search and memcached running on a real-system Intel heterogeneous prototype demonstrates that Octopus-Man improves energy efficiency by up to 41% (CPU power) and up to 15% (system power) over an all-brawny WSC design while adhering to specified QoS targets.

As applications such as Apple Siri, Google Now, Microsoft Cortana, and Amazon Echo continue to gain traction, web-service companies are adopting large deep neural networks (DNN) for machine learning challenges such as image processing, speech recognition, natural language processing, among others. A number of open questions arise as to the design of a server platform specialized for DNN and how modern warehouse scale computers (WSCs) should be outfitted to provide DNN as a service for these applications.

Rampant dynamism due to load fluctuations, co runner changes, and varying levels of interference poses a threat to application quality of service (QoS) and has limited our ability to allow co-locations in modern warehouse scale computers (WSCs). Instruction set features such as the non-temporal memory access hints found in modern ISAs (both ARM and x86) may be useful in mitigating these effects. However, despite the challenge of this dynamism and the availability of an instruction set mechanism that might help address the problem, a key capability missing in the system software stack in modern WSCs is the ability to dynamically transform (and re-transform) the executing application code to apply these instruction set features when necessary. In this work we introduce protean code, a novel approach for enacting arbitrary compiler transformations at runtime for native programs running on commodity hardware with negligible (<;1%) overhead. The fundamental insight behind the underlying mechanism of protean code is that, instead of maintaining full control throughout the program's execution as with traditional dynamic optimizers, protean code allows the original binary to execute continuously and diverts control flow only at a set of virtualized points, allowing rapid and seamless rerouting to the new code variants. In addition, the protean code compiler embeds IR with high-level semantic information into the program, empowering the dynamic compiler to perform rich analysis and transformations online with little overhead. Using a fully functional protean code compiler and runtime built on LLVM, we design PC3D, Protean Code for Cache Contention in Datacenters. PC3D dynamically employs non-temporal access hints to achieve utilization improvements of up to 2.8x (1.5x on average) higher than state-of-the-art contention mitigation runtime techniques at a QoS target of 98%.

One of the key challenges for improving efficiency in warehouse scale computers (WSCs) is to improve server utilization while guaranteeing the quality of service (QoS) of latency-sensitive applications. To this end, prior work has proposed techniques to precisely predict performance and QoS interference to identify 'safe' application co-locations. However, such techniques are only applicable to resources shared across cores. Achieving such precise interference prediction on real-system simultaneous multithreading (SMT) architectures has been a significantly challenging open problem due to the complexity introduced by sharing resources within a core. In this paper, we demonstrate through a real-system investigation that the fundamental difference between resource sharing behaviors on CMP and SMT architectures calls for a redesign of the way we model interference. For SMT servers, the interference on different shared resources, including private caches, memory ports, as well as integer and floating-point functional units, do not correlate with each other. This insight suggests the necessity of decoupling interference into multiple resource sharing dimensions. In this work, we propose SMiTe, a methodology that enables precise performance prediction for SMT co-location on real-system commodity processors. With a set of Rulers, which are carefully designed software stressors that apply pressure to a multidimensional space of shared resources, we quantify application sensitivity and contentiousness in a decoupled manner. We then establish a regression model to combine the sensitivity and contentiousness in different dimensions to predict performance interference. Using this methodology, we are able to precisely predict the performance interference in SMT co-location with an average error of 2.80% on SPEC CPU2006 and 1.79% on Cloud Suite. Our evaluation shows that SMiTe allows us to improve the utilization of WSCs by up to 42.57% while enforcing an application's QoS requirements.

This work presents an end-to-end methodology for quantifying the performance and power benefits of simultaneous multithreading (SMT) for HPC centers and applies this methodology to a production system and workload. Ultimately, SMT’s value system-wide depends on whether users effectively employ SMT at the application level. However, predicting SMT’s benefit for HPC applications is challenging; by doubling the number of threads, the application’s characteristics may change. This work proposes statistical modeling techniques to predict the speedup SMT confers to HPC applications. This approach, accurate to within 8%, uses only lightweight, transparent performance monitors collected during a single run of the application.

Accelerators are becoming prevalent in high performance computing as a way of achieving increased computational capacity within a smaller power budget. Effectively utilizing the raw compute capacity made available by these systems, however, remains a challenge because it can require a substantial investment of programmer time to port and optimize code to effectively use novel accelerator hardware. In this paper we present a methodology for isolating and modeling the performance of common performance-critical patterns of code (so-called idioms) and other relevant behavioral characteristics from large scale HPC applications which are likely to perform favorably on Intel Xeon Phi. The benefits of the methodology are twofold: (1) it directs programmer efforts toward the regions of code most likely to benefit from porting to the Xeon Phi and (2) provides speedup estimates for porting those regions of code. We then apply the methodology to the stencil idiom, showing performance improvements of up to a factor of 4.7× on stencil-based benchmark codes.

Suppose one is considering purchase of a computer equipped with accelerators. Or suppose one has access to such a computer and is considering porting code to take advantage of the accelerators. Is there a reason to suppose the purchase cost or programmer effort will be worth it? It would be nice to able to estimate the expected improvements in advance of paying money or time. We exhibit an analytical framework and tool-set for providing such estimates: the tools first look for user-defined idioms that are patterns of computation and data access identified in advance as possibly being able to benefit from accelerator hardware. A performance model is then applied to estimate how much faster these idioms would be if they were ported and run on the accelerators, and a recommendation is made as to whether or not each idiom is worth the porting effort to put them on the accelerator and an estimate is provided of what the overall application speedup would be if this were done.

Binary instrumentation facilitates the insertion of additional code into an executable in order to observe or modify the executable's behavior. There are two main approaches to binary instrumentation: static and dynamic binary instrumentation. In this paper we present a static binary instrumentation toolkit for Linux on the x86/x86_64 platforms, PEBIL (PMaC's Efficient Binary Instrumentation Toolkit for Linux). PEBIL is similar to other toolkits in terms of how additional code is inserted into the executable. However, it is designed with the primary goal of producing efficient-running instrumented code. To this end, PEBIL uses function level code relocation in order to insert large but fast control structures. Furthermore, the PEBIL API provides tool developers with the means to insert lightweight hand-coded assembly rather than relying solely on the insertion of instrumentation functions. These features enable the implementation of efficient instrumentation tools with PEBIL. The overhead introduced for basic block counting by PEBIL is an average of 65% of the overhead of Dyninst, 41% of the overhead of Pin, 15% of the overhead of DynamoRIO, and 8% of the overhead of Valgrind.

SPECFEM3D_GLOBE is a spectral-element application enabling the simulation of global seismic wave propagation in 3D anelastic, anisotropic, rotating and self-gravitating Earth models at unprecedented resolution. A fundamental challenge in global seismology is to model the propagation of waves with periods between 1 and 2 seconds, the highest frequency signals that can propagate clear across the Earth. These waves help reveal the 3D structure of the Earth's deep interior and can be compared to seismographic recordings. We broke the 2 second barrier using the 62K processor Ranger system at TACC. Indeed we broke the barrier using just half of Ranger, by reaching a period of 1.84 seconds with sustained 28.7 Tflops on 32K processors. We obtained similar results on the XT4 Franklin system at NERSC and the XT4 Kraken system at University of Tennessee Knoxville, while a similar run on the 28K processor Jaguar system at ORNL, which has better memory bandwidth per processor, sustained 35.7 Tflops (a higher flops rate) with a 1.94 shortest period.

A systematic study of the effects of complexity of prediction methodology on its accuracy for a set of real applications on a variety of HPC systems is performed. Results indicate that the use of any single, simple synthetic metric to predict performance does an inadequate job, and the use of a linear combination of these simple metrics with optimized weights also performs poorly. Better, however, are methodologies that rely on the convolution of an application "transfer function" based on tracing information with system performance data measured by simple benchmarks. This latter methodology can predict performance with an average accuracy of 80%, based on the current work.