Voltage and Temperature-Aware SSTA Using Neural Network Delay Model

With the emergence of voltage scaling as one of the most powerful power reduction techniques, it has been important to support voltage scalable statistical static timing analysis (SSTA) in deep submicrometer process nodes. In this paper, we propose a single delay model of logic gate using neural network which comprehensively captures process, voltage, and temperature variation along with input slew and output load. The number of simulation programs with integrated circuit emphasis (SPICE) required to create this model over a large voltage and temperature range is found to be modest and 4x less than that required for a conventional table-based approach with comparable accuracy. We show how the model can be used to derive sensitivities required for linear SSTA for an arbitrary voltage and temperature. Our experimentation on ISCAS 85 benchmarks across a voltage range of 0.9-1.1V shows that the average error in mean delay is less than 1.08% and average error in standard deviation is less than 2.85%. The errors in predicting the 99% and 1% probability point are 1.31% and 1%, respectively, with respect to SPICE. The two potential applications of voltage-aware SSTA have been presented, i.e., one for improving the accuracy of timing analysis by considering instance-specific voltage drops in power grids and the other for determining optimum supply voltage for target yield for dynamic voltage scaling applications.

Improving Superscalar Instruction Dispatch And Issue By Exploiting Dynamic Code Sequences

Superscalar processors currently have the potential to fetch multiple basic blocks per cycle by employing one of several recently proposed instruction fetch mechanisms. However, this increased fetch bandwidth cannot be exploited unless pipeline stages further downstream correspondingly improve. In particular,register renaming a large number of instructions per cycle is diDcult. A large instruction window, needed to receive multiple basic blocks per cycle, will slow down dependence resolution and instruction issue. This paper addresses these and related issues by proposing (i) partitioning of the instruction window into multiple blocks, each holding a dynamic code sequence; (ii) logical partitioning of the registerjle into a global file and several local jles, the latter holding registers local to a dynamic code sequence; (iii) the dynamic recording and reuse of register renaming information for registers local to a dynamic code sequence. Performance studies show these mechanisms improve performance over traditional superscalar processors by factors ranging from 1.5 to a little over 3 for the SPEC Integer programs. Next, it is observed that several of the loops in the benchmarks display vector-like behavior during execution, even if the static loop bodies are likely complex for compile-time vectorization. A dynamic loop vectorization mechanism that builds on top of the above mechanisms is briefly outlined. The mechanism vectorizes up to 60% of the dynamic instructions for some programs, albeit the average number of iterations per loop is quite small.

Variable Granularity Access Tracking Scheme for Improving the Performance of Software Transactional Memory

Software transactional memory (STM) has been proposed as a promising programming paradigm for shared memory multi-threaded programs as an alternative to conventional lock based synchronization primitives. Typical STM implementations employ a conflict detection scheme, which works with uniform access granularity, tracking shared data accesses either at word/cache line or at object level. It is well known that a single fixed access tracking granularity cannot meet the conflicting goals of reducing false conflicts without impacting concurrency adversely. A fine grained granularity while improving concurrency can have an adverse impact on performance due to lock aliasing, lock validation overheads, and additional cache pressure. On the other hand, a coarse grained granularity can impact performance due to reduced concurrency. Thus, in general, a fixed or uniform granularity access tracking (UGAT) scheme is application-unaware and rarely matches the access patterns of individual application or parts of an application, leading to sub-optimal performance for different parts of the application(s). In order to mitigate the disadvantages associated with UGAT scheme, we propose a Variable Granularity Access Tracking (VGAT) scheme in this paper. We propose a compiler based approach wherein the compiler uses inter-procedural whole program static analysis to select the access tracking granularity for different shared data structures of the application based on the application's data access pattern. We describe our prototype VGAT scheme, using TL2 as our STM implementation. Our experimental results reveal that VGAT-STM scheme can improve the application performance of STAMP benchmarks from 1.87% to up to 21.2%.

Enhancing Speedup in Network Processing Applications by Exploiting Instruction Reuse with Flow Aggregation

Instruction reuse is a microarchitectural technique that improves the execution time of a program by removing redundant computations at run-time. Although this is the job of an optimizing compiler, they do not succeed many a time due to limited knowledge of run-time data. In this paper we examine instruction reuse of integer ALU and load instructions in network processing applications. Specifically, this paper attempts to answer the following questions: (1) How much of instruction reuse is inherent in network processing applications?, (2) Can reuse be improved by reducing interference in the reuse buffer?, (3) What characteristics of network applications can be exploited to improve reuse?, and (4) What is the effect of reuse on resource contention and memory accesses? We propose an aggregation scheme that combines the high-level concept of network traffic i.e. "flows" with a low level microarchitectural feature of programs i.e. repetition of instructions and data along with an architecture that exploits temporal locality in incoming packet data to improve reuse. We find that for the benchmarks considered, 1% to 50% of instructions are reused while the speedup achieved varies between 1% and 24%. As a side effect, instruction reuse reduces memory traffic and can therefore be considered as a scheme for low power.

Making STMs Cache Friendly with Compiler Transformations

Software transactional memory (STM) is a promising programming paradigm for shared memory multithreaded programs. In order for STMs to be adopted widely for performance critical software, understanding and improving the cache performance of applications running on STM becomes increasingly crucial, as the performance gap between processor and memory continues to grow. In this paper, we present the most detailed experimental evaluation to date, of the cache behavior of STM applications and quantify the impact of the different STM factors on the cache misses experienced by the applications. We find that STMs are not cache friendly, with the data cache stall cycles contributing to more than 50% of the execution cycles in a majority of the benchmarks. We find that on an average, misses occurring inside the STM account for 62% of total data cache miss latency cycles experienced by the applications and the cache performance is impacted adversely due to certain inherent characteristics of the STM itself. The above observations motivate us to propose a set of specific compiler transformations targeted at making the STMs cache friendly. We find that STM's fine grained and application unaware locking is a major contributor to its poor cache behavior. Hence we propose selective Lock Data co-location (LDC) and Redundant Lock Access Removal (RLAR) to address the lock access misses. We find that even transactions that are completely disjoint access parallel, suffer from costly coherence misses caused by the centralized global time stamp updates and hence we propose the Selective Per-Partition Time Stamp (SPTS) transformation to address this. We show that our transformations are effective in improving the cache behavior of STM applications by reducing the data cache miss latency by 20.15% to 37.14% and improving execution time by 18.32% to 33.12% in five of the 8 STAMP applications.

Combining heterogeneous classifiers for relational databases

Practical usage of machine learning is gaining strategic importance in enterprises looking for business intelligence. However, most enterprise data is distributed in multiple relational databases with expert-designed schema. Using traditional single-table machine learning techniques over such data not only incur a computational penalty for converting to a flat form (mega-join), even the human-specified semantic information present in the relations is lost. In this paper, we present a practical, two-phase hierarchical meta-classification algorithm for relational databases with a semantic divide and conquer approach. We propose a recursive, prediction aggregation technique over heterogeneous classifiers applied on individual database tables. The proposed algorithm was evaluated on three diverse datasets. namely TPCH, PKDD and UCI benchmarks and showed considerable reduction in classification time without any loss of prediction accuracy. (C) 2012 Elsevier Ltd. All rights reserved.

Exploiting the Structure of the Constraint Graph for Efficient Points-to Analysis

Points-to analysis is a key compiler analysis. Several memory related optimizations use points-to information to improve their effectiveness. Points-to analysis is performed by building a constraint graph of pointer variables and dynamically updating it to propagate more and more points-to information across its subset edges. So far, the structure of the constraint graph has been only trivially exploited for efficient propagation of information, e.g., in identifying cyclic components or to propagate information in topological order. We perform a careful study of its structure and propose a new inclusion-based flow-insensitive context-sensitive points-to analysis algorithm based on the notion of dominant pointers. We also propose a new kind of pointer-equivalence based on dominant pointers which provides significantly more opportunities for reducing the number of pointers tracked during the analysis. Based on this hitherto unexplored form of pointer-equivalence, we develop a new context-sensitive flow-insensitive points-to analysis algorithm which uses incremental dominator update to efficiently compute points-to information. Using a large suite of programs consisting of SPEC 2000 benchmarks and five large open source programs we show that our points-to analysis is 88% faster than BDD-based Lazy Cycle Detection and 2x faster than Deep Propagation. We argue that our approach of detecting dominator-based pointer-equivalence is a key to improve points-to analysis efficiency.

Implications of program phase behavior on timing analysis

Knowledge about program worst case execution time (WCET) is essential in validating real-time systems and helps in effective scheduling. One popular approach used in industry is to measure execution time of program components on the target architecture and combine them using static analysis of the program. Measurements need to be taken in the least intrusive way in order to avoid affecting accuracy of estimated WCET. Several programs exhibit phase behavior, wherein program dynamic execution is observed to be composed of phases. Each phase being distinct from the other, exhibits homogeneous behavior with respect to cycles per instruction (CPI), data cache misses etc. In this paper, we show that phase behavior has important implications on timing analysis. We make use of the homogeneity of a phase to reduce instrumentation overhead at the same time ensuring that accuracy of WCET is not largely affected. We propose a model for estimating WCET using static worst case instruction counts of individual phases and a function of measured average CPI. We describe a WCET analyzer built on this model which targets two different architectures. The WCET analyzer is observed to give safe estimates for most benchmarks considered in this paper. The tightness of the WCET estimates are observed to be improved for most benchmarks compared to Chronos, a well known static WCET analyzer.

Relative roles of instruction count and cycles per instruction in WCET estimation

Most of the existing WCET estimation methods directly estimate execution time, ET, in cycles. We propose to study ET as a product of two factors, ET = IC * CPI, where IC is instruction count and CPI is cycles per instruction. Considering directly the estimation of ET may lead to a highly pessimistic estimate since implicitly these methods may be using worst case IC and worst case CPI. We hypothesize that there exists a functional relationship between CPI and IC such that CPI=f(IC). This is ascertained by computing the covariance matrix and studying the scatter plots of CPI versus IC. IC and CPI values are obtained by running benchmarks with a large number of inputs using the cycle accurate architectural simulator, Simplescalar on two different architectures. It is shown that the benchmarks can be grouped into different classes based on the CPI versus IC relationship. For some benchmarks like FFT, FIR etc., both IC and CPI are almost a constant irrespective of the input. There are other benchmarks that exhibit a direct or an inverse relationship between CPI and IC. In such a case, one can predict CPI for a given IC as CPI=f(IC). We derive the theoretical worst case IC for a program, denoted as SWIC, using integer linear programming(ILP) and estimate WCET as SWIC*f(SWIC). However, if CPI decreases sharply with IC then measured maximum cycles is observed to be a better estimate. For certain other benchmarks, it is observed that the CPI versus IC relationship is either random or CPI remains constant with varying IC. In such cases, WCET is estimated as the product of SWIC and measured maximum CPI. It is observed that use of the proposed method results in tighter WCET estimates than Chronos, a static WCET analyzer, for most benchmarks for the two architectures considered in this paper.

Prioritizing constraint evaluation for efficient points-to analysis

Pervasive use of pointers in large-scale real-world applications continues to make points-to analysis an important optimization-enabler. Rapid growth of software systems demands a scalable pointer analysis algorithm. A typical inclusion-based points-to analysis iteratively evaluates constraints and computes a points-to solution until a fixpoint. In each iteration, (i) points-to information is propagated across directed edges in a constraint graph G and (ii) more edges are added by processing the points-to constraints. We observe that prioritizing the order in which the information is processed within each of the above two steps can lead to efficient execution of the points-to analysis. While earlier work in the literature focuses only on the propagation order, we argue that the other dimension, that is, prioritizing the constraint processing, can lead to even higher improvements on how fast the fixpoint of the points-to algorithm is reached. This becomes especially important as we prove that finding an optimal sequence for processing the points-to constraints is NP-Complete. The prioritization scheme proposed in this paper is general enough to be applied to any of the existing points-to analyses. Using the prioritization framework developed in this paper, we implement prioritized versions of Andersen's analysis, Deep Propagation, Hardekopf and Lin's Lazy Cycle Detection and Bloom Filter based points-to analysis. In each case, we report significant improvements in the analysis times (33%, 47%, 44%, 20% respectively) as well as the memory requirements for a large suite of programs, including SPEC 2000 benchmarks and five large open source programs.

NUcache: an efficient multicore cache organization based on next-use distance

The effectiveness of the last-level shared cache is crucial to the performance of a multi-core system. In this paper, we observe and make use of the DelinquentPC - Next-Use characteristic to improve shared cache performance. We propose a new PC-centric cache organization, NUcache, for the shared last level cache of multi-cores. NUcache logically partitions the associative ways of a cache set into MainWays and DeliWays. While all lines have access to the MainWays, only lines brought in by a subset of delinquent PCs, selected by a PC selection mechanism, are allowed to enter the DeliWays. The PC selection mechanism is an intelligent cost-benefit analysis based algorithm that utilizes Next-Use information to select the set of PCs that can maximize the hits experienced in DeliWays. Performance evaluation reveals that NUcache improves the performance over a baseline design by 9.6%, 30% and 33% respectively for dual, quad and eight core workloads comprised of SPEC benchmarks. We also show that NUcache is more effective than other well-known cache-partitioning algorithms.

Extended histories: improving regularity and performance in correlation prefetchers

Data Prefetchers identify and make use of any regularity present in the history/training stream to predict future references and prefetch them into the cache. The training information used is typically the primary misses seen at a particular cache level, which is a filtered version of the accesses seen by the cache. In this work we demonstrate that extending the training information to include secondary misses and hits along with primary misses helps improve the performance of prefetchers. In addition to empirical evaluation, we use the information theoretic metric entropy, to quantify the regularity present in extended histories. Entropy measurements indicate that extended histories are more regular than the default primary miss only training stream. Entropy measurements also help corroborate our empirical findings. With extended histories, further benefits can be achieved by triggering prefetches during secondary misses also. In this paper we explore the design space of extended prefetch histories and alternative prefetch trigger points for delta correlation prefetchers. We observe that different prefetch schemes benefit to a different extent with extended histories and alternative trigger points. Also the best performing design point varies on a per-benchmark basis. To meet these requirements, we propose a simple adaptive scheme that identifies the best performing design point for a benchmark-prefetcher combination at runtime. In SPEC2000 benchmarks, using all the L2 accesses as history for prefetcher improves the performance in terms of both IPC and misses reduced over techniques that use only primary misses as history. The adaptive scheme improves the performance of CZone prefetcher over Baseline by 4.6% on an average. These performance gains are accompanied by a moderate reduction in the memory traffic requirements.

Loop transformations: convexity, pruning, and optimization

High-level loop transformations are a key instrument in mapping computational kernels to effectively exploit the resources in modern processor architectures. Nevertheless, selecting required compositions of loop transformations to achieve this remains a significantly challenging task; current compilers may be off by orders of magnitude in performance compared to hand-optimized programs. To address this fundamental challenge, we first present a convex characterization of all distinct, semantics-preserving, multidimensional affine transformations. We then bring together algebraic, algorithmic, and performance analysis results to design a tractable optimization algorithm over this highly expressive space. Our framework has been implemented and validated experimentally on a representative set of benchmarks running on state-of-the-art multi-core platforms.

Automatic compilation of MATLAB programs for synergistic execution on heterogeneous processors

MATLAB is an array language, initially popular for rapid prototyping, but is now being increasingly used to develop production code for numerical and scientific applications. Typical MATLAB programs have abundant data parallelism. These programs also have control flow dominated scalar regions that have an impact on the program's execution time. Today's computer systems have tremendous computing power in the form of traditional CPU cores and throughput oriented accelerators such as graphics processing units(GPUs). Thus, an approach that maps the control flow dominated regions to the CPU and the data parallel regions to the GPU can significantly improve program performance. In this paper, we present the design and implementation of MEGHA, a compiler that automatically compiles MATLAB programs to enable synergistic execution on heterogeneous processors. Our solution is fully automated and does not require programmer input for identifying data parallel regions. We propose a set of compiler optimizations tailored for MATLAB. Our compiler identifies data parallel regions of the program and composes them into kernels. The problem of combining statements into kernels is formulated as a constrained graph clustering problem. Heuristics are presented to map identified kernels to either the CPU or GPU so that kernel execution on the CPU and the GPU happens synergistically and the amount of data transfer needed is minimized. In order to ensure required data movement for dependencies across basic blocks, we propose a data flow analysis and edge splitting strategy. Thus our compiler automatically handles composition of kernels, mapping of kernels to CPU and GPU, scheduling and insertion of required data transfer. The proposed compiler was implemented and experimental evaluation using a set of MATLAB benchmarks shows that our approach achieves a geometric mean speedup of 19.8X for data parallel benchmarks over native execution of MATLAB.

Runtime dependence computation and execution of loops on heterogeneous systems

GPUs have been used for parallel execution of DOALL loops. However, loops with indirect array references can potentially cause cross iteration dependences which are hard to detect using existing compilation techniques. Applications with such loops cannot easily use the GPU and hence do not benefit from the tremendous compute capabilities of GPUs. In this paper, we present an algorithm to compute at runtime the cross iteration dependences in such loops. The algorithm uses both the CPU and the GPU to compute the dependences. Specifically, it effectively uses the compute capabilities of the GPU to quickly collect the memory accesses performed by the iterations by executing the slice functions generated for the indirect array accesses. Using the dependence information, the loop iterations are levelized such that each level contains independent iterations which can be executed in parallel. Another interesting aspect of the proposed solution is that it pipelines the dependence computation of the future level with the actual computation of the current level to effectively utilize the resources available in the GPU. We use NVIDIA Tesla C2070 to evaluate our implementation using benchmarks from Polybench suite and some synthetic benchmarks. Our experiments show that the proposed technique can achieve an average speedup of 6.4x on loops with a reasonable number of cross iteration dependences.