An Accurate and Efficient Time-Division Parallelization of Cycle Accurate Architectural Simulators

There is wide agreement that one of the most important impediments to the performance of current and future pipelined superscalar processors is the presence of conditional branches in the instruction stream. Speculative execution seems to be one solution of choice to the branch problem, but speculative work is discarded if a branch is mispredicted. Therefore, we need a very accurate branch predictor; 95% accuracy is not good enough. This paper proposes branch classification to help improve the accuracy of branch predictors. Branch classification allows an individual branch instruction to be associated with the branch predictor best suited to predict its direction. Using this approach, a hybrid branch predictor can be constructed such that each component branch predictor predicts those branches for which it is best suited. This paper suggests one classification scheme, analyzes several branch predictors, and proposes a hybrid branch predictor that achieves higher prediction accuracy than any branch predictor previously reported in the literature.

The schedulability analysis of real-time embedded systems requires worst case execution time (WCET) analysis for the individual tasks. Bounding WCET involves not only language-level program path analysis, but also modeling the performance impact of complex micro-architectural features present in modern processors. In this paper, we statically analyze the execution time of embedded software on processors with speculative execution. The speculation of conditional branch outcomes (branch prediction) significantly improves a program's execution time. Thus, accurate modeling of control speculation is important for calculating tight WCET estimates. We present a parameterized framework to model the different branch prediction schemes. We further consider the complex interaction between speculative execution and instruction cache performance, that is, the fact that speculatively executed blocks can generate additional cache hits/misses. We extend our modeling to capture this effect of branch prediction on cache performance. Starting with the control flow graph of a program, our technique uses integer linear programming to estimate the program's WCET. The accuracy of our method is demonstrated by tight estimates obtained on realistic benchmarks.

The multicore revolution and the ever-increasing complexity of computing systems is dramatically changing sys-tem design, analysis and programming of computing platforms. Future architectures will feature hundreds to thousands of simple processors and on-chip memories connected through a network-on-chip. Architectural simulators will remain primary tools for design space exploration, software development and performance evaluation of these massively parallel architectures. However, architectural simulation performance is a serious concern, as virtual platforms and simulation technology are not able to tackle the complexity of thousands of core future scenarios. The main contribution of this paper is the development of a new simulation approach and technology for many core processors which exploit the enormous parallel processing capability of low-cost and widely available General Purpose Graphic Processing Units (GPGPU). The simulation of many-core architectures exhibits indeed a high level of parallelism and is inherently parallelizable, but GPGPU acceleration of architectural simulation requires an in-depth revision of the data structures and functional partitioning traditionally used in parallel simulation. We demonstrate our GPGPU simulator on a target architecture composed by several cores (i.e. ARM ISA based), with instruction and data caches, connected through a Network-on-Chip (NoC). Our experiments confirm the feasibility of our approach.

Performance simulation of software for Multiprocessor System-on-a-Chips (MPSoC) suffers from poor tool support. Cycle accurate simulation at Instruction Set Simulation level is too slow and inefficient for any design of realistic size. Behavioral simulation, though useful for functional analysis at high level, does not provide any performance information that is crucial for design and analysis of MPSoC implementations. As a consequence, designers are often reduced to manually annotate performance information onto behavioral models, which contributes further to inefficiency and inaccuracy. In this paper, we use structural performance models to provide fast and accurate simulation of software for MPSoC. We generate structural models automatically using GCC with accurate performance annotation while considering optimizations for instruction selection, branch prediction, and pipeline interlock. Our structural models are able to simulate at several orders of magnitude faster than ISS and provide less than 1% error on performance estimation. These models allow realistic MPSoC design space explorations based on performance characteristics with simulation speed comparable to behavioral simulation. We validate our simulation models with several benchmarks and demonstrate our approach with a design case study of an MPEG-2 decoder.

Current software-based microarchitecture simulators are many orders of magnitude slower than the hardware they simulate. Hence, most microarchitecture design studies draw their conclusions from drastically truncated benchmark simulations that are often inaccurate and misleading. The Sampling Microarchitecture Simulation (SMARTS) framework is an approach to enable fast and accurate performance measurements of full-length benchmarks. SMARTS accelerates simulation by selectively measuring in detail only an appropriate benchmark subset. SMARTS prescribes a statistically sound procedure for configuring a systematic sampling simulation run to achieve a desired quantifiable confidence in estimates. Analysis of the SPEC CPU2000 benchmark suite shows that CPI can be estimated to within /spl plusmn/3% with 99.7% confidence by measuring fewer than 50 million instructions per benchmark. In practice, inaccuracy in microarchitectural state initialization introduces an additional uncertainty which we empirically bound to /spl sim/2% for the tested benchmarks. We present two implementations of SMARTS that both achieve an average error of only 0.64% on CPI. SMARTSim constructs accurate model state through functional warming - continuously warming large microarchitectural structures (e.g., caches and the branch predictor) while functionally simulating the billions of instructions between measurements - reducing average simulation turnaround from 5.5 days to 7.0 hours. TurboSMARTSim replaces functional warming with live-points - checkpoints that store a bare minimum of functionally-warmed state for accurate simulation of a limited execution window - further reducing average turnaround to 91 seconds.

Special issue on <i>specialized</i> computer architecture simulators that see the present and may hold the future

In the push for ever increasing performance out of processor architectures, there is a need to expand beyond the limitations of existing scalar approaches. Superscalar architectures provide one such means. By dynamically executing more than one instruction per clock cycle, superscalar architectures can improve performance without relying solely on technology improvements for these gains. This paper examines a new technique for superscalar control implementation, called the schedule table. The schedule table facilitates dependency checking, out of order instruction issue, out of order execution, branch prediction, speculative execution, precise interrupts, and fast and efficient misprediction recovery.

This is a study of the formal verification of a VLIW microprocessor that imitates the Intel Itanium [9][12][17] in features such as predicated execution, register remapping, advanced and speculative loads, and branch prediction. The formal verification is done with the Burch and Dill flushing technique [5] by exploiting the properties of Positive Equality [3][4]. The contributions include an extensive use of conservative approximations in abstracting portions of the processor and a framework for decomposition of the Boolean evaluation of the correctness formula. The conservative approximations are applied automatically when abstracting a memory whose forwarding logic is not affected by stalling conditions that preserve the correctness of the memory semantics for the same memory. These techniques allow a reduction of more than a factor of 4 in the CPU time for the formal verification of the most complex processor model examined relative to the monolithic evaluation of the correctness formula for a version of the same processor where conservative approximations are not applied.

Due to the prohibitive cost of worst-case timing analysis for modern processors, the design of time-predictable processors has become increasingly important for hard real-time and safety-critical systems. However, to the best of our knowledge currently there is no effective and widely accepted metric to quantitatively evaluate time predictability of processors, which greatly impedes the advancement of time-predictable processor design. This paper first introduces the concept of architectural time predictability (ATP), which separates timing uncertainty concerns caused by hardware from software. We then propose a metric called Architectural Time-predictability Factor (ATF) to measure architectural time predictability. Our evaluation on a Very Long Instruction Word (VLIW) processor indicates that ATF is an effective metric to quantitatively evaluate architectural time predictability of a whole processor as well as its architectural and microarchitectural components such as caches, branch prediction, speculative execution, parallel pipelines, and Scratch-Pad Memory (SPM). Thus ATF can be used to quantitatively guide architectural design for enhancing time predictability or making better trade-offs between performance and time predictability.

It has been shown that control flow decisions often reduce the amount of fine-grain, or instruction-level parallelism in many computations. There exists various architectural solutions to this problem, such as branch prediction and speculative execution. Here we present an alternative solution, based on the use of partial evaluation for statically increasing the parallelism of a computation. In particular we explore the capability of a partial evaluator to remove control flow decisions.A partial evaluator is described, which specializes data flow graphs produced by a compiler for a fine-grained processor array. The data flow graphs can be viewed as the equivalent of a machine-code program for conventional processor, and consequently some aspects of the architecture can be taken into account by the partial evaluator.Results from the use of the partial evaluator show that it can significantly improve the performance of a computation. The performance improvement is partly due to improved parallelism, but mostly due to a reduce dynamic instruction count, i.e. a reduction in the number of operations executed.

Architectural simulation is time-consuming, and the trend towards hundreds of cores is making sequential simulation even slower. Existing parallel simulation techniques either scale poorly due to excessive synchronization, or sacrifice accuracy by allowing event reordering and using simplistic contention models. As a result, most researchers use sequential simulators and model small-scale systems with 16-32 cores. With 100-core chips already available, developing simulators that scale to thousands of cores is crucial. We present three novel techniques that, together, make thousand-core simulation practical. First, we speed up detailed core models (including OOO cores) with instruction-driven timing models that leverage dynamic binary translation. Second, we introduce bound-weave, a two-phase parallelization technique that scales parallel simulation on multicore hosts efficiently with minimal loss of accuracy. Third, we implement lightweight user-level virtualization to support complex workloads, including multiprogrammed, client-server, and managed-runtime applications, without the need for full-system simulation, sidestepping the lack of scalable OSs and ISAs that support thousands of cores. We use these techniques to build zsim, a fast, scalable, and accurate simulator. On a 16-core host, zsim models a 1024-core chip at speeds of up to 1,500 MIPS using simple cores and up to 300 MIPS using detailed OOO cores, 2-3 orders of magnitude faster than existing parallel simulators. Simulator performance scales well with both the number of modeled cores and the number of host cores. We validate zsim against a real Westmere system on a wide variety of workloads, and find performance and microarchitectural events to be within a narrow range of the real system.

Emerging massively parallel architectures such as a general-purpose processor plus many-core programmable accelerators are creating an increasing demand for novel methods to perform their architectural simulation. Most state-of-the-art simulation technologies are exceedingly slow and the need to model full system many-core architectures adds further to the complexity issues. This paper presents a fast, scalable and parallel simulator, which uses a novel methodology to accelerate the simulation of a many-core coprocessor using GPU platforms. The main idea is to use. The target architecture of the associated . Simulation of many target nodes is mapped to the many hardware-threads available on highly parallel GPU platforms. This paper presents a novel methodology to accelerate the simulation of many-core coprocessors using GPU platforms. We demonstrate the challenges, feasibility and benefits of our idea to use heterogeneous system (CPU and GPU) to simulate future architecture of many-core heterogeneous platforms. The target architecture selected to evaluate our methodology consists of an ARM general purpose CPU coupled withmany-core coprocessor with thousands of simple in-order cores connected in a tile network. This work presents optimization techniques used to parallelize the simulation specifically for acceleration on GPUs. We partition the full system simulation between CPU and GPU, where the target general purpose CPU is simulated on the host CPU, whereas the many-core coprocessor is simulated on the NVIDIA Tesla 2070 GPU platform. Our experiments show performance of up to 50 MIPS when simulating the entire heterogeneous chip, and high scalability with increasing cores on coprocessor.

Speculative execution is an optimization technique used in modern processors by which predicted instructions are executed in advance with an objective of overlapping the latencies of slow operations. Branch prediction and load value speculation are examples of speculative execution used in modern pipelined processors to avoid execution stalls. However, speculative executions incur a performance penalty as an execution rollback when there is a misprediction. In this work, we propose to aid speculative execution with approximate computing by relaxing the execution rollback penalty associated with a misprediction. We propose a sensitivity analysis method for data and branches in a program to identify the data load and branch instructions that can be executed without any rollback in the pipeline and yet can ensure a certain user-specified quality of service of the application with a probabilistic reliability. Our analysis is based on statistical methods, particularly hypothesis testing and Bayesian analysis. We perform an architectural simulation of our proposed approximate execution and report the benefits in terms of CPU cycles and energy utilization on selected applications from the AxBench, ACCEPT, and Parsec 3.0 benchmarks suite.

The reliance on simulation for the design of a new architecture continues to increase as the complexity of the architecture increases with more cores integrated and a complex communication fabric. A good compromise between the simulation speed and accuracy should be made in order to explore the wide architectural design space at the early design phase. In this work, we present a novel manycore simulator that achieves fast and scalable performance with a trace-driven parallel simulation technique. Since the communication delay affects the simulation accuracy heavily, the proposed simulator supports a high-speed accurate on-chip network model, which can capture link/buffer congestions. Its development is motivated by an industry project to design a new manycore architecture with 96 cores with a NoC architecture. Experimental results show that the proposed simulator could run real parallel benchmark applications with more than 450 Kcycles/sec/core. It marks a Pareto-optimal simulator between speed and accuracy, compared with existent simulators.

A conditional branch predictor based on weightless neural networks