Search

Emerging persistent memory (PM) technologies promise the performance of DRAM with the durability of Flash. Several language-level persistency models have emerged recently to aid programming recoverable data structures in PM. Unfortunately, these persistency models are built upon hardware primitives that impose stricter ordering constraints on PM operations than the persistency models require. Alternative solutions use fixed and inflexible hardware logging techniques to relax ordering constraints on PM operations, but do not readily apply to general synchronization primitives employed by language-level persistency models. Instead, we propose StrandWeaver, a hardware strand persistency model, to minimally constrain ordering on PM operations. StrandWeaver manages PM order within a strand, a logically independent sequence of operations within a thread. PM operations that lie on separate strands are unordered and may drain concurrently to PM. StrandWeaver implements primitives under strand persistency to allow programmers to improve concurrency and relax ordering constraints on updates as they drain to PM. Furthermore, we design mechanisms that map persistency semantics in high-level language persistency models to the primitives implemented by StrandWeaver. We demonstrate that StrandWeaver can enable greater concurrency of PM operations than existing ISA-level ordering mechanisms, improving performance by up to $1.97 \times (1.45 \times avg.)$.

The commercial release of byte-addressable persistent memories, such as Intel/Micron 3D XPoint memory, is imminent. Ongoing research has sought mechanisms to allow programmers to implement recoverable data structures in these new main memories. Ensuring recoverability requires programmer control of the order of persistent stores; recent work proposes persistency models as an extension to memory consistency to specify such ordering. Prior work has considered persistency models at the abstraction of the instruction set architecture. Instead, we argue for extending the language-level memory model to provide guarantees on the order of persistent writes. We explore a taxonomy of guarantees a language-level persistency model might provide, considering both atomicity and ordering constraints on groups of persistent stores. Then, we propose and evaluate Acquire-Release Persistency (ARP), a language-level persistency model for C++11. We describe how to compile code written for ARP to a state-of-the-art ISA-level persistency model. We then consider enhancements to the ISA-level persistency model that can distinguish memory consistency constraints required for proper synchronization but unnecessary for correct recovery. With these optimizations, we show that ARP increases performance by up to 33.2% (19.8% avg.) over coding directly to the baseline ISA-level persistency model for a suite of persistent-write-intensive workloads.

Modern processors use branch prediction as an optimization to improve processor performance. Predictors have become larger and increasingly more sophisticated in order to achieve higher accuracies which are needed in high performance cores. However, branch prediction can also be a source of side channel exploits, as one context can deliberately change the branch predictor state and alter the instruction flow of another context. Current mitigation techniques either sacrifice performance for security, or fail to guarantee isolation when retaining the accuracy. Achieving both has proven to be challenging.In this work we address this by, (1) introducing the notions of steady-state and transient branch predictor accuracy, and (2) showing that current predictors increase their misprediction rate by as much as 90% on average when forced to flush branch prediction state to remain secure. To solve this, (3) we introduce the branch retention buffer, a novel mechanism that partitions only the most useful branch predictor components to isolate separate contexts. Our mechanism makes thread isolation practical, as it stops the predictor from executing cold with little if any added area and no warm-up overheads. At the same time our results show that, compared to the state-of-the-art, average misprediction rates are reduced by 15-20% without increasing area, leading to a 2% performance increase.

Full-system simulation frameworks such as gem5 are used extensively to evaluate research ideas and for design-space exploration. Moreover, energy-efficiency has become the key design constraint in recent years and many works use a separate power modelling framework to evaluate energy consumption. While such tools are convenient and flexible, they are known to contain sources of error which are often not fully understood and potentially impact the conclusions drawn from investigations. This work enables accurate, hardware-validated performance, power, and energy modelling of CPUs by first presenting a methodology to evaluate and identify sources of error in CPU performance models, and secondly developing empirical power models optimised for use with such performance models. Hierarchical clustering, correlation analysis, and regression techniques are used to identify sources of error without requiring detailed CPU specifications and enable existing models to be improved, new models to be developed, validation of simulator changes, and testing of model suitability for specific use-cases. Furthermore, the GemStone open-source software tool is presented, which automates the process of characterising hardware platforms, identifying sources of error in gem5 models, applying power analysis, and quantifying the effect of errors on the performance, power, and energy estimations. In addition, the mean percentage error in execution time was found to swing from −51% to +10% between two versions of the same gem5 model, underlining the need for an automated tool to validate models against reference hardware, ensuring accuracy and consistency.

Heterogeneous multi-processors are designed to bridge the gap between performance and energy efficiency in modern embedded systems. This is achieved by pairing Out-of-Order (OoO) cores, yielding performance through aggressive speculation and latency masking, with In-Order (InO) cores, that preserve energy through simpler design. By leveraging migrations between them, workloads can therefore select the best setting for any given energy/delay envelope. However, migrations introduce execution overheads that can hurt performance if they happen too frequently. Finding the optimal migration frequency is critical to maximize energy savings while maintaining acceptable performance. We develop a simulation methodology that can 1) isolate the hardware effects of migrations from the software, 2) directly compare the performance of different core types, 3) quantify the performance degradation and 4) calculate the cost of migrations for each case. To showcase our methodology we run mibench, a microbenchmark suite, and show that migrations can happen as fast as every 100k instructions with little performance loss. We also show that, contrary to numerous recent studies, hypothetical designs do not need to share all of their internal components to be able to migrate at that frequency. Instead, we propose a feasible system that shares level 2 caches and a translation lookaside buffer that matches performance and efficiency. Our results show that there are phases comprising up to 10% that a migration to the OoO core leads to performance benefits without any additional energy cost when running on the InO core, and up to 6% of phases where a migration to the InO core can save energy without affecting performance. When considering a policy that focuses on improving the energy-delay product, results show that on average 66% of the phases can be migrated to deliver equal or better system operation without having to aggressively share the entire memory system or to revert to migration periods finer than 100k instructions.

When analyzing a distributed computer system, we often observe that the complex interplay among processor, node, and network sub-systems can profoundly affect the performance and power efficiency of the distributed computer system. Therefore, to effectively cross-optimize hardware and software components of a distributed computer system, we need a full-system simulation infrastructure that can precisely capture the complex interplay. Responding to the aforementioned need, we present dist-gem5, a flexible, detailed, and open-source full-system simulation infrastructure that can model and simulate a distributed computer system using multiple simulation hosts. Then we validate dist-gem5 against a physical cluster and show that the latency and bandwidth of the simulated network sub-system are within 18% of the physical one. Compared with the single threaded and parallel versions of gem5, dist-gem5 speeds up the simulation of a 63-node computer cluster by 83.1x and 12.8x, respectively.

Accurate and stable CPU power modelling is fundamental in modern system-on-chips (SoCs) for two main reasons: 1) they enable significant online energy savings by providing a run-time manager with reliable power consumption data for controlling CPU energy-saving techniques; 2) they can be used as accurate and trusted reference models for system design and exploration. We begin by showing the limitations in typical performance monitoring counter (PMC) based power modelling approaches and illustrate how an improved model formulation results in a more stable model that efficiently captures relationships between the input variables and the power consumption. Using this as a solid foundation, we present a methodology for adding thermal-awareness and analytically decomposing the power into its constituting parts. We develop and validate our methodology using data recorded from a quad-core ARM Cortex-A15 mobile CPU and we achieve an average prediction error of 3.7% across 39 diverse workloads, 8 Dynamic Voltage-Frequency Scaling (DVFS) levels and with a CPU temperature ranging from 31 degrees C to 91 degrees C. Moreover, we measure the effect of switching cores offline and decompose the existing power model to estimate the static power of each CPU and L2 cache, the dynamic power due to constant background (BG) switching, and the dynamic power caused by the activity of each CPU individually. Finally, we provide our model equations and software tools for implementing in a run-time manager or for using with an architectural simulator, such as gem5.

Simulation tools are indispensable to computer architects. Detailed execution-driven CPU models offer high accuracy, but at the cost of simulation speed. Trace-driven simulation is widely adopted to alleviate this problem, especially for studies focusing on memory-system exploration. Ideally, trace-driven core models will mimic out-of-order processors executing full-system workloads to enable computer architects to evaluate modern systems. Additionally, to be useful to the broader community the tracing and replay models should be publicly available. However, existing trace-driven approaches are limited in their applicability and availability. We propose elastic traces in which we accurately capture data and load/store order dependencies by instrumenting a detailed out-of-order processor model. In contrast to existing work, we do not rely on offline analysis of timestamps, and instead use accurate dependency information tracked inside the processor pipeline. We thereby account for the effects of speculation and branch misprediction resulting in a more accurate trace playback. We provide a trace player that honours the dependencies and thus adapts its execution time to memory-system changes, as would the actual CPU. Compared to the detailed CPU, our trace player achieves a speed-up of 6–8 times. When modifying the memory-system parameters, the average error in absolute execution time is 7% for SPEC 2006 benchmarks on a bare metal system and 17% for HPC benchmarks on Linux. Relative performance is predicted with less than 3% error, achieving fast and accurate system performance exploration. We make this functionality available to the broader community via a widely-used open source full-system simulator.

As computer systems become increasingly complex, the need for fast and accurate simulation tools increases. Accurate but slow processor core models are often substituted with simple trace players to achieve faster memory-system simulation. However, existing trace-driven simulation techniques are limited in their applicability and availability. In this work, we capture elastic traces containing out-of-order core dependencies and effects of speculative execution, which overcome limitations of existing work. Additionally, we make our capture and replay modelling available in the gem5 simulator. Our trace-driven CPU achieves a speed-up of 6–8x compared to the reference core and predicts the performance with less than 1% error on average when the memory-system is changed.

Advanced Synchronization Facility (ASF) is an AMD64 hardware extension for lock-free data structures and transactional memory. It provides a speculative region that atomically executes speculative accesses in the region. Five new instructions are added to demarcate the region, use speculative accesses selectively, and control the speculative hardware context. Programmers can use speculative regions to build flexible multi-word atomic primitives with no additional software support by relying on the minimum guarantee of available ASF hardware resources for lock-free programming. Transactional programs with high-level TM language constructs can either be compiled directly to the ASF code or be linked to software TM systems that use ASF to accelerate transactional execution. In this paper we develop an out-of-order hardware design to implement ASF on a future AMD processor and evaluate it with an in-house simulator. The experimental results show that the combined use of the L1 cache and the LS unit is very helpful for the performance robustness of ASF-based lock free data structures, and that the selective use of speculative accesses enables transactional programs to scale with limited ASF hardware resources.

The adoption of multi- and many-core architectures for mainstream computing undoubtedly brings profound changes in the way software is developed. In particular, the use of fine grained locking as the multi-core programmer’s coordination methodology is considered by more and more experts as a dead-end. The transactional memory (TM) programming paradigm is a strong contender to become the approach of choice for replacing locks and implementing atomic operations in concurrent programming. Combining sequences of concurrent operations into atomic transactions allows a great reduction in the complexity of both programming and verification, by making parts of the code appear to execute sequentially without the need to program using fine-grained locking. Transactions remove from the programmer the burden of figuring out the interaction among concurrent operations that happen to conflict when accessing the same locations in memory. The EU-funded FP7 VELOX project designs, implements and evaluates an integrated TM stack, spanning from programming language to the hardware support, and including runtime and libraries, compilers, and application environments. This paper presents an overview of the VELOX TM stack and its associated challenges and contributions.

AMD's Advanced Synchronization Facility (ASF) is an x86 instruction set extension proposal intended to simplify and speed up the synchronization of concurrent programs. In this paper, we report our experiences using ASF for implementing transactional memory. We have extended a C/C++ compiler to support language-level transactions and generate code that takes advantage of ASF. We use a software fallback mechanism for transactions that cannot be committed within ASF (e.g., because of hardware capacity limitations). Our evaluation uses a cycle-accurate x86 simulator that we have extended with ASF support. Building a complete ASF-based software stack allows us to evaluate the performance gains that a user-level program can obtain from ASF. Our measurements on a wide range of benchmarks indicate that the overheads traditionally associated with software transactional memories can be significantly reduced with the help of ASF.

Memory persistency models enable maintaining recoverable data structures in persistent memories and prior work has proposed ISA-level persistency models. In addition to these models, we argue for extending language-level memory models to provide persistence semantics. We present a taxonomy of guarantees a language-level persistency model could provide and characterize their programmability and performance.