John P. VanDyke

Topics on measuring real power usage on high performance computing platforms

Power has recently been recognized as one of the major obstacles in fielding a Peta-FLOPs class system. To reach Exa-FLOPs, the challenge will certainly be compounded. In this paper we will discuss a number of High Performance Computing power related topics. We first describe our implementation of a scalable power measurement framework that has enabled us to examine real power use (current draw). [Using this framework, samples were obtained at a per-node (socket) granularity, at frequencies of up to 100 samples per second.] Additionally, we describe how we applied this capability to implement power conserving measures on our Catamount Light Weight Kernel, where we achieved an 80% improvement. This ability has enabled us to quantify the amount of energy used by applications and to contrast application energy use between a Light Weight and General Purpose operating system. Finally, we show application energy use increases proportionally with the increase in run-time due to operating system noise. Areas of future interest will also be discussed.

3-D prestack Kirchhoff depth migration: From prototype to production in a massively parallel processor environment

Portable, production-scale 3-D prestack Kirchhoff depth migration software capable of full-volume imaging has been successfully implemented and applied to a six-million trace (46.9 Gbyte) marine data set from a salt/subsalt play in the Gulf of Mexico. Velocity model building and updates use an image-driven strategy and were performed in a Sun Sparc environment. Images obtained by 3-D prestack migration after three velocity iterations are substantially better focused and reveal drilling targets that were not visible in images obtained from conventional 3-D poststack time migration. Amplitudes are well preserved, so anomalies associated with known reservoirs conform to the petrophysical predictions. Prototype development was on an 8-node Intel iPSC860 computer; the production version was run on an 1824-node Intel Paragon computer. The code has been successfully ported to CRAY (T3D) and Unix workstation (PVM) environments.

Energy Delay Product

In this chapter, data from both the CPU frequency tuning experiments (Chap. 6) and the network bandwidth experiments (Chap. 7) are analyzed using a range of fused metrics based on Energy Delay Product (EDP). The analysis in this chapter demonstrates how multiple metrics can be combined and observed as a single fused metric. Additionally, a form of weighted EDP is used to more highly prioritize, or weight, performance over energy savings.

Report of experiments and evidence for ASC L2 milestone 4467 : demonstration of a legacy application's path to exascale.

This report documents thirteen of Sandias contributions to the Computational Systems and Software Environment (CSSE) within the Advanced Simulation and Computing (ASC) program between fiscal years 2009 and 2012. It describes their impact on ASC applications. Most contributions are implemented in lower software levels allowing for application improvement without source code changes. Improvements are identified in such areas as reduced run time, characterizing power usage, and Input/Output (I/O). Other experiments are more forward looking, demonstrating potential bottlenecks using mini-application versions of the legacy codes and simulating their network activity on Exascale-class hardware. The purpose of this report is to prove that the team has completed milestone 4467-Demonstration of a Legacy Applications Path to Exascale. Cielo is expected to be the last capability system on which existing ASC codes can run without significant modifications. This assertion will be tested to determine where the breaking point is for an existing highly scalable application. The goal is to stretch the performance boundaries of the application by applying recent CSSE RD in areas such as resilience, power, I/O, visualization services, SMARTMAP, lightweight LWKs, virtualization, simulation, and feedback loops. Dedicated system time reservations and/or CCC allocations will be used to quantify the impact of system-level changes to extend the life and performance of the ASC code base. Finally, a simulation of anticipated exascale-class hardware will be performed using SST to supplement the calculations. Determine where the breaking point is for an existing highly scalable application: Chapter 15 presented the CSSE work that sought to identify the breaking point in two ASC legacy applications-Charon and CTH. Their mini-app versions were also employed to complete the task. There is no single breaking point as more than one issue was found with the two codes. The results were that applications can expect to encounter performance issues related to the computing environment, system software, and algorithms. Careful profiling of runtime performance will be needed to identify the source of an issue, in strong combination with knowledge of system software and application source code.

Investigating An API for Resilient Exascale Computing

Increased HPC capability comes with increased complexity, part counts, and fault occurrences. In- creasing the resilience of systems and applications to faults is a critical requirement facing the viability of exascale systems, as the overhead of traditional checkpoint/restart is projected to outweigh its bene ts due to fault rates outpacing I/O bandwidths. As faults occur and propagate throughout hardware and software layers, pervasive noti cation and handling mechanisms are necessary. This report describes an initial investigation of fault types and programming interfaces to mitigate them. Proof-of-concept APIs are presented for the frequent and important cases of memory errors and node failures, and a strategy proposed for lesystem failures. These involve changes to the operating system, runtime, I/O library, and application layers. While a single API for fault handling among hardware and OS and application system-wide remains elusive, the e ort increased our understanding of both the mountainous challenges and the promising trailheads. 3

Tuning CPU Power During Application Runtime

In Chap. 5, the focus was on reducing power by exploiting idle cycles. In the experiments outlined in this chapter and in Chap. 7 the focus is on identifying opportunities to reduce the energy use of a running application without affecting performance. As mentioned previously, determining what is, and is not, an acceptable trade-off between energy and performance is somewhat subjective. The motivation of the following experiments is to show that significant energy savings can be achieved by tuning architectural components, specifically the CPU, on a per application basis during execution.

Network Bandwidth Tuning During Application Runtime

Chapter 6 showed the effects that CPU frequency tuning had on application energy and performance. In this chapter, we will evaluate the impact that tuning the network bandwidth has on energy and performance of real scientific computing applications running at large scale. This analysis will evaluate the impact on total node energy, in contrast to the CPU frequency tuning experiments that focused on CPU energy. These experiments provide further evidence that tuning components on large-scale High Performance Computing (HPC) platforms can result in significant energy savings.

3-D Prestack Kirchhoff Depth Migration: From Prototype to Production In a MPP Environment.

Portable, production-scale 3-D prestack Kirchhoff migration software capable of full-volume imaging has been successfully implemented and applied to a six million trace (46.9 Gbyte) data set from a salt/subsalt play in the Gulf of Mexico. Velocity model building and updates use an imagedriven strategy and were performed in a Sun Sparc environment. Images obtained after three velocity iterations are substantially better focused and reveal drilling targets that were not visible in images obtained from conventional 3-D postsk the production version was run on an 1824 node Intel Paragon. The code has been successfully ported to CRAY (T3D), and Parallel Virtual Machine (PVM) environments.

Massively parallel computing and the mid-course tracking problem

No abstract available

High‐Temperature Expansions for Classical Systems

We have derived series in powers of K to 8‐th order for the two‐point function, susceptibility and free energy of the model Hamiltonian, ΣR⇒ {W[Q(R⇒)] + K/2 Σδ⇒ Q(R⇒)·Q(R⇒+δ⇒)}, where Q is a real tensor, and W is an arbitrary, even function of Q. Special cases include the spin‐S Ising model and the classical Heisenberg model, for both of which W = 0. Additional applications include models for structural phase transitions, tricritical points, He3‐He4 mixtures, liquid crystals, and the class of models employed in renormalization — group studies of critical phenomena. Detailed evaluation of susceptibility series have been carried out on the standard 2, 3, and 4‐dimensional lattices. The series for the scalar case on the triangular net is presented herein.