scholarly journals The loop fusion for data localization

2020 ◽  
Vol 11 (3) ◽  
pp. 17-31
Author(s):  
Борис Яковлевич Штейнберг ◽  
Олег Борисович Штейнберг ◽  
Александр Александрович Василенко
Keyword(s):  
Loop Fusion

Для улучшения локализации данных используется слияние циклов. Слияние циклов, имеющих общие переменные, может ускорить исполнение за счёт уменьшения количества кэш-промахов. Это преобразование известно давно, но компиляторы выполняют его лишь для простейших случаев. Наши улучшенные алгоритмы используют предварительные преобразования для корректного слияния циклов, имеющих разное количество итераций и информационные зависимости.

Energy minimization with loop fusion and multi-functional-unit scheduling for multidimensional DSP

2008 ◽  
Vol 68 (4) ◽  
pp. 443-455 ◽  
Author(s):  
Meikang Qiu ◽  
Edwin H.-M. Sha ◽  
Meilin Liu ◽  
Man Lin ◽  
Shaoxiong Hua ◽  
...  
Keyword(s):  
Energy Minimization ◽  
Functional Unit ◽  
Loop Fusion

scholarly journals THEMIS: COMPONENT DEPENDENCE METADATA IN ADAPTIVE PARALLEL APPLICATIONS

2001 ◽  
Vol 11 (04) ◽  
pp. 455-470 ◽  
Author(s):  
PAUL H. J. KELLY ◽  
OLAV BECKMANN ◽  
TONY FIELD ◽  
SCOTT B. BADEN
Keyword(s):  
Performance Optimization ◽  
Programming Model ◽  
Data Distribution ◽  
Parallel Applications ◽  
Problem Instance ◽  
Placement Optimization ◽  
Loop Fusion ◽  
Run Time ◽  
Optimization Loop ◽  
Execution Order

This paper describes THEMIS, a programming model and run-time library being designed to support cross-component performance optimization through explicit manipulation of the computation's iteration space at run-time. Each component is augmented with "component dependence metadata", which characterizes the constraints on its execution order, data distribution and memory access order. We show how this supports dynamic adaptation of each component to exploit the available resources, the context in which its operands are generated, and results are used, and the evolution of the problem instance. Using a computational fluid dynamics visualization example as motivation, we show how component dependence metadata provides a framework in which a number of interesting optimizations become possible. Examples include data placement optimization, loop fusion, tiling, memoization, checkpointing and incrementalization.

scholarly journals Optimal Design Based on Closed-Loop Fusion for Velocity Bandwidth Expansion of Optical Target Tracking System

Sensors ◽  
2019 ◽  
Vol 19 (1) ◽  
pp. 133 ◽  
Author(s):  
Yao Mao ◽  
Wei Ren ◽  
Yong Luo ◽  
Zhijun Li
Keyword(s):  
Optimal Design ◽  
Target Tracking ◽  
Sensor Fusion ◽  
Closed Loop ◽  
Tracking System ◽  
Micro Electro Mechanical System ◽  
Measurement Range ◽  
Improve Measurement ◽  
Loop Fusion ◽  
Bandwidth Expansion

Micro-electro-mechanical system (MEMS) gyro is one of the extensively used inertia sensors in the field of optical target tracking (OTT). However, velocity closed-loop bandwidth of the OTT system is limited due to the resonance and measurement range issues of MEMS gyro. In this paper, the generalized sensor fusion framework, named the closed-loop fusion (CLF), is analyzed, and the optimal design principle of filter is proposed in detail in order to improve measurement of the bandwidth of MEMS gyro by integrating information of MEMS accelerometers. The fusion error optimization problem, which is the core issue of fusion design, can be solved better through the feedback compensation law of CLF framework and fusion filter optimal design. Differently from conventional methods, the fusion filter of CLF can be simply and accurately designed, and the determination of superposition of fusion information can also be effectively avoided. To show the validity of the proposed method, both sensor fusion simulations and closed-loop experiments of optical target tracking system have yielded excellent results.

Sign in / Sign up

Export Citation Format

Share Document