Dynamic profiling and trace cache generation

Dynamic program optimization is increasingly important for achieving good runtime performance. A key issue is how to select which code to optimize. One approach is to dynamically detect traces, long sequences of instructions spanning multiple methods, which are likely to execute to completion. Traces are easy to optimize and have been shown to be a good unit for optimization. The paper reports on a new approach for dynamically detecting, creating and storing traces in a Java virtual machine. We first describe four important criteria for a successful trace strategy: good instruction stream coverage, low dispatch rate, cache stability, and optimizability of traces. We then present our approach based on branch correlation graphs. A branch correlation graph stores information about the correlation between pairs of branches, as well as additional state information. We present the complete design for an efficient implementation of the system, including a detailed discussion of the trace cache and profiling mechanisms. We have implemented an experimental framework to measure the traces generated by our approach in a direct-threaded Java VM (SableVM) and we present experimental results to show that the traces we generate meet the design criteria.

For the last decade, the Java Virtual Machine (JVM) has been a popular platform to host languages other than Java. Language implementation frameworks like Truffle allow the implementation of dynamic languages such as JavaScript or Ruby with competitive performance and completeness. However, statically typed languages are still rare under Truffle. We present Sulong, an LLVM IR interpreter that brings all LLVM-based languages including C, C++, and Fortran in one stroke to the JVM. Executing these languages on the JVM enables a wide area of future research, including high-performance interoperability between high-level and low-level languages, combination of static and dynamic optimizations, and a memory-safe execution of otherwise unsafe and unmanaged languages.

This study takes the North District of Yaxinyuan in Shijiazhuang as a case study and explores the design strategies for the renovation of old residential areas based on the theory of urban micro-renewal. The research finds that old residential areas in Shijiazhuang generally face problems such as building aging, lack of public service facilities, and insufficient greening, while the North District of Yaxinyuan has particularly prominent contradictions in aspects like building safety, aging-friendly facilities, and the vitality of public spaces. Through questionnaire surveys and on-site investigations, the authors propose an "acupuncture-style" minimally invasive renovation strategy. This study constructs a renovation method system of "collaborative governance of multiple subjects—coupling of policies and community needs—dynamic program optimization", providing a "low-cost, low-intervention, and high-inclusiveness" renovation model for the transformation of old residential areas in northern industrial cities. It holds reference value for promoting the modernization of urban governance and the improvement of the quality of human settlements.

The trace cache is a recently proposed solution to achieving high instruction fetch bandwidth by buffering and reusing dynamic instruction traces. This work presents a new block-based trace cache implementation that can achieve higher IPC performance with more efficient storage of traces. Instead of explicitly storing instructions of a trace, pointers to blocks constituting a trace are stored in a much smaller trace table. The block-based trace cache renames fetch addresses at the basic block level and stores aligned blocks in a block cache. Traces are constructed by accessing the replicated block cache using block pointers from the trace table. Performance potential of the block-based trace cache is quantified and compared with perfect branch prediction and perfect fetch schemes. Comparing to the conventional trace cache, the block-based design can achieve higher IPC, with less impact on cycle time.Results: Using the SPECint95 benchmarks, a 16-wide realistic design of a block-based trace cache can improve performance 75% over a baseline design and to within 7% of a baseline design with perfect branch prediction. With idealized trace prediction, it is shown the block-based trace cache with an 1K-entry block cache achieves the same performance of the conventional trace cache with 32K entries.

Research on energy management strategy for fuel cell hybrid electric vehicles based on improved dynamic programming and air supply optimization

As process technology continues to shrink, a large number of bitcells in on-chip caches is expected to be faulty. The number of defective cells varies from die-to-die, wafer-to-wafer, and in the field of application depends on the run-time operating conditions (e.g., supply voltage and frequency). Those trends necessitate i) to study fault-tolerant (FT) cache mechanisms in a wide spectrum of fault-probabilities and ii) to devise appropriate FT techniques that must be able to adapt their FT capacity to the volume of defective locations of the target faulty caches. It is well known that keeping the cache capacity, block size and the volume of defective cells constant, the average number of misses, due to faulty cells, decreases as the associativity of the cache increases. To this end, we propose DARCA, a Defect-Aware Reconfigurable Cache Architecture, which is equipped with the ability of dynamically varying its associativity according to the volume of defective cells. To keep the hardware overhead very small, as the associativity of the cache is multiplied by a power of two, its block size is divided by the same number. Since almost all contemporary processors use prefetching, we also applied DARCA to prefetch-assisted caches. By performing cycle-accurate simulations for the SPEC2006 benchmarks assuming a wide range of fault-probabilities, we show that DARCA compares favorably against several known FT cache mechanisms with respect to the performance loss caused by defective cells.

Dynamic program optimization offers performance improvements far beyond those possible with traditional compile-time optimization. These gains are due to the ability to exploit both architectural and input data set characteristics that are unknown prior to execution time. In this paper, we propose a novel framework for dynamic program optimization, ADAPT (Automated De-coupled Adaptive Program Transformation), that builds on the strengths of existing approaches. The key to our framework is the de-coupling of the dynamic compilation of new code variants from the dynamic selection of these variants at their points of use. This allows code generation to occur concurrently with program execution, removing dynamic compilation overheads from the critical path. We present a compilation system, based on the Polaris optimizing compiler, that automatically applies this framework to general plugged-in optimization techniques. We evaluate our system on three programs from the SPEC floating point benchmark suite by dynamically applying loop distribution, loop unrolling, loop tiling and automatic parallelization. We show that our techniques can improve performance by as much as 70% over statically optimized code.

This paper presents a method based on dynamic programming optimization to design distribution and industrial networks with unbalanced customers. Low-voltage and medium-voltage network customers can be three-phase (i.e., industrial and commercial) and single-phase (e.g,. residential and rural areas). Moreover, American medium-voltage distribution networks have single-phase lines (rural areas) and three-phase lines (urban and primary feeders). Optimization of an unbalanced network implies optimal assignation of single-phase customers to each phase. This works considers different conductor types, tapering, power losses in lines, capacity and voltage drop constraints and deterministic loads, all employing single-phase and three-phase lines. The necessary modifications to apply it by single-wire earth return systems is presented.

A dynamic programming optimization algorithm has been applied on a transit bus model in MATLAB in order to assess the fuel economy improvement potential by implementing a hydraulic hybrid powertrain system. The numerical model parameters have been calibrated using experimental data obtained on a Belgrade?s public transport bus. This experiment also provided the representative driving cycle on which to conduct simulation analyses. Various functional parameters of a hydraulic hybrid system have been evaluated for obtaining the best possible fuel economy. Dynamic programming optimization runs have been completed for various hydraulic accumulator sizes, preload values and accumulator foam quantities. It has been shown that a fuel economy improvement of 28% can be achieved by implementing such a system.

This paper extends previously published results on the application of a dynamic programming optimization approach to the planning of systems, with special emphasis on the long-term expansion of power systems. Future uncertainties about loads, equipment costs, etc. are considered explicitly in the method described. A practical computational solution to the resulting stochastic optimization is obtained by means of the recently developed open-loop feedback approximation. Further, it is shown how various known approaches of production cost simulation can be used in conjunction with the dynamic programming optimization to accommodate large systems that may be represented with great technical detail.

Reservoir optimal scheduling is a typical multi-constrained,dynamic,non-linear optimization problem.To solve this problem,a Dynamic Programming-Particle Swarm Optimization (DP-PSO) algorithm was used to solving.This algorithm used the optimal Dynamic Programming (DP) principle to convert the reservoir optimal scheduling problem to multistage decision-making sub-problems;the solution of each sub-problem was got by particle swarm optimization algorithm.The numerical experiments show that with more time in calculation,the reliability of the DP-PSO is superior to the general DP algorithm,and the calculation time of DP-PSO is less than DP-Genetic Algorithm (DP-GA).

The Irrigation Area (D. I.) Mambal, which passes through Badung Regency, Denpasar City, and Tabanan Regency, is the largest irrigation water supplied by the Ayung River, covering an area of 5.963 Ha. Despite the Ayung River’s substantial water potential, the D. I. Mambal experiences water shortages during certain months. This research aims to improve the effectiveness and efficiency of irrigation water use based on the Global Planting Management Plan (RTTG) using simulation methods and dynamic program optimization. Simulations were carried out under low conditions, normal and sufficient dependable discharges, using both existing and alternative RTTG. The objective function of the dynamic optimization seeks to maximize revenue gain from the applied RTTG. The existing cropping pattern at the beginning of planting in October showed an average proportion of fulfillment of water irrigation needs at 85%. Under the Alternative I condition, with planting beginning in November, the average proportion of fulfillment of irrigation water needs was 89%. In Alternative II conditions, with planting beginning in December, the average proportion of fulfillment of irrigation water needs was 87%. By optimizing the water discharge using the dynamic program, the irrigation profit for the existing cropping pattern (October) amounted to IDR 491,816,154,938. The highest profit was obtained using the Alternative II cropping pattern (December), totaling IDR 606,675,369,830. Meanwhile, the lowest profit was obtained in the Alternative I cropping pattern (November), which was IDR 360,767,292,361. The analysis showed that the Alternative II cropping pattern, starting with the first rice planting period in December, yields the most optimal results. The analysis considers the optimized air allocation and irrigation benefits obtained from the third cropping pattern.

Live migration of Java Virtual Machines (JVMs) consumes significant amounts of time and resources, imposing relevant application performance overhead. This problem is specially hard when memory modified by applications changes faster than it can be transferred through the network (to a remote host). Current solutions to this problem resort to several techniques which depend on high-speed networks and application throttling, require lots of CPU time to compress memory, or need explicit assistance from the application. We propose a novel approach, Garbage Collector (GC) assisted JVM Live Migration for Java Server Applications (ALMA). ALMA makes a snapshot to be migrated containing a minimal amount of application state, by taking into account the amount of reachable memory (i.e. live data) detected by the GC. The main novelty of ALMA is the following: ALMA analyzes the JVM heap looking for regions in which a collection phase is advantageous w.r.t. the network bandwidth available (i.e. it pays to collect because a significant amount of memory will not be part of the snapshot). ALMA is implemented on OpenJDK 8 and extends CRIU (a Linux disk-based process checkpoint/restore tool) to support process live migration over the network. We evaluate ALMA using well-known JVM performance benchmarks (SPECjvm2008 and DaCapo), and by comparing it to other previous approaches. ALMA shows very good performance results.

It is well known that hybrid electric vehicles (HEVs) have fewer requirements to transmission due to the favorable performance of electric motor for traction. However, multi-gear transmission with more gears is still beneficial in improving acceleration and fuel economy performance of the vehicles than one with fewer gears, even not as much as in conventional vehicles. On the other hand, use of multi-gear transmission with more gears inevitably adds complexity to vehicle structure and control, consequently resulting in high cost. In this paper, the sensitivity of vehicle performance to transmission gears for a parallel HEV has been analyzed by using SOC based Dynamic Programming. The results generated in this paper provide helpful guidance for HEV design.

Conventional slope stability analysis uses various Limit Equilibrium (LE) methods to determine the minimum factor of safety and its associated critical failure mechanism. These methods frequently assume that collapse will follow a particular assumed geometry, which are effective for simple geotechnical problems, yet may encounter difficulties when considering complex problems. For such complex problems, one effective solution lies in the use of Limit Analysis (LA) based on the upper-bound of plasticity in conjunction with a discretization procedure known as Discontinuity Layout Optimization (DLO), which can be an effective means of determining a critical failure mechanism without the limitation of an assumed slip surface. This paper compares the use of LE (Spencer Method with dynamic programming optimization) and LA for several well-known examples of complex slopes. It is shown LA generally provides slightly lower factors of safety than rigorous LE and manages to handle complexities more effectively without assuming the geometry of the slip surface. It is demonstrated that using the DLO with LA can be applied to various extreme value problems beyond classical slope problems.