High Performance Reconfigurable Computing Research Articles

Towards Complete and Scalable Emulation of Quantum Algorithms on High-Performance Reconfigurable Computers

Contemporary quantum computers face many critical challenges that limit their usefulness for practical applications. A primary limiting factor is classical-to-quantum (C2Q) data encoding, which requires specific circuits for quantum state initialization. The required state initialization circuits are often complex and violate decoherence constraints, particularly for I/O intensive applications. Existing Noisy Intermediate-Scale Quantum (NISQ) devices are noise-sensitive and have low quantum bit (qubit) counts, thus limiting the applicability of C2Q circuits for encoding large and realistic datasets. This has made the study of complete and realistic circuits that include data encoding challenging and has also led to a heavy dependency on costly and resource-intensive simulations on classical platforms. In this work, we propose a cost-effective, classical-hardware-accelerated framework for realistic and complete emulation of quantum algorithms. The emulation framework incorporates components for the critical C2Q data encoding process, as well as architectures for quantum algorithms such as the quantum Haar transform (QHT). The framework is used to investigate optimizations for C2Q and QHT algorithms, and the corresponding optimized quantum circuits are presented. The framework is implemented on a High-Performance Reconfigurable Computer (HPRC) which emulates the proposed QHT circuits combined with proposed C2Q data encoding methods. For performance benchmarks, CPU-based emulations and simulations on a state-of-the-art quantum computing simulator are also carried out. Results show that the proposed hardware-accelerated emulation framework is more efficient in terms of speed and scalability compared to CPU-based emulation and simulation.

Quantum Dimension Reduction for Pattern Recognition in High-Resolution Spatio-Spectral Data

The promises of advanced quantum computing technology have driven research in the simulation of quantum computers on classical hardware, where the feasibility of quantum algorithms for real-world problems can be investigated. In domains such as High Energy Physics (HEP) and Remote Sensing Hyperspectral Imagery, classical computing systems are held back by enormous readouts of high-resolution data. Due to the multi-dimensionality of the readout data, processing and performing pattern recognition operations for this enormous data are both computationally intensive and time-consuming. In this article, we propose a methodology that utilizes Quantum Haar Transform (QHT) and a modified Grover's search algorithm for time-efficient dimension reduction and dynamic pattern recognition in data sets that are characterized by high spatial resolution and high dimensionality. QHT is performed on the data to reduce its dimensionality at preserved spatial locality, while the modified Grover's search algorithm is used to search for dynamically changing multiple patterns in the reduced data set. By performing search operations on the reduced data set, processing overheads are minimized. Moreover, quantum techniques produce results in less time than classical dimension reduction and search methods. The feasibility of the proposed methodology is verified by emulating the quantum algorithms on classical hardware based on field programmable gate arrays (FPGAs). We present designs of the quantum circuits for multi-dimensional QHT and multi-pattern Grover's search. We also present two emulation techniques and the corresponding hardware architectures for this methodology. A high performance reconfigurable computer (HPRC) was used for the experimental evaluation, and high-resolution images were used as the input data set. Analysis of the methods and implications of the experimental results are discussed.

Efficient Computation Techniques and Hardware Architectures for Unitary Transformations in Support of Quantum Algorithm Emulation

As the development of quantum computers progresses rapidly, continuous research efforts are ongoing for simulation and emulation of quantum algorithms on classical platforms. Software simulations require use of large-scale, costly, and resource-hungry supercomputers, while hardware emulators make use of fast Field-Programmable-Gate-Array (FPGA) accelerators, but are limited in accuracy and scalability. This work presents a cost-effective FPGA-based emulation platform that demonstrates improved scalability, accuracy, and throughput compared to existing FPGA-based emulators. In this work, speed and area trade-offs between different proposed emulation architectures and computation techniques are investigated. For example, stream-based computation is proposed that greatly reduces resource utilization, improves system scalability in terms of the number of emulated quantum bits, and allows for dynamically changing algorithm inputs. The proposed techniques assume that the unitary transformation of the quantum algorithm is known, and the matrix values can be pre-computed or generated dynamically. 32-bit floating-point precision is used for high accuracy and the architectures are fully pipelined to ensure high throughput. As case studies for emulation, the quantum Fourier transform and Grover’s search algorithms are investigated and quantum circuits for multi-pattern Grover’s search are also proposed. Experimental evaluation and analysis of the emulation architectures and computation techniques are provided for the investigated quantum algorithms. The emulation framework is prototyped on a high-performance reconfigurable computing (HPRC) system and the results show quantitative improvement over existing FPGA-based emulators.

In-FPGA Instrumentation Framework for OpenCL-Based Designs

The productivity achieved when developing applications on high-performance reconfigurable heterogeneous computing (HPRHC) systems is increased by using the Open Computing Language (OpenCL). However, the hardware produced by OpenCL compilers in field-programmable gate arrays (FPGAs) can result in severe performance bottlenecks that are challenging to solve. The problem is compounded by the fact that the generated netlist details are disorganized, making them mostly unreadable and only partially visible to designers. This paper proposes an in-FPGA instrumentation method and a new framework for extracting the FPGA-cycle-accurate timing performances of OpenCL-based designs. The results clearly show that the chosen execution model for OpenCL-based designs strongly affects the timing performance when it is not properly implemented. Our framework is implemented on an HPRHC platform that contains a CPU and two Arria10 FPGAs, and it is evaluated with a wide variety of benchmarks with different complexities. After testing on the reported benchmarks, the average logic overhead for one inserted instrument is 0.2 % of the total amount of adaptive look-up tables (ALUTs) and 0.1 % of the total registers in an FPGA. This resource utilization is between 1.5 and six times lower than those reported in the best previously published works. The scalability of the framework is also evaluated by inserting up to 50 instruments. The experimental results show that the average logic utilization per instrument is 0.19 % of the ALUTs and 0.17 % of the registers in the FPGA when 50 instruments are inserted.

Dimension Reduction Using Quantum Wavelet Transform on a High-Performance Reconfigurable Computer

The high resolution of multidimensional space-time measurements and enormity of data readout counts in applications such as particle tracking in high-energy physics (HEP) is becoming nowadays a major challenge. In this work, we propose combining dimension reduction techniques with quantum information processing for application in domains that generate large volumes of data such as HEP. More specifically, we propose using quantum wavelet transform (QWT) to reduce the dimensionality of high spatial resolution data. The quantum wavelet transform takes advantage of the principles of quantum mechanics to achieve reductions in computation time while processing exponentially larger amount of information. We develop simpler and optimized emulation architectures than what has been previously reported, to perform quantum wavelet transform on high-resolution data. We also implement the inverse quantum wavelet transform (IQWT) to accurately reconstruct the data without any losses. The algorithms are prototyped on an FPGA-based quantum emulator that supports double-precision floating-point computations. Experimental work has been performed using high-resolution image data on a state-of-the-art multinode high-performance reconfigurable computer. The experimental results show that the proposed concepts represent a feasible approach to reducing dimensionality of high spatial resolution data generated by applications such as particle tracking in high-energy physics.

Análisis de Eficiencia en Sistemas de Cómputo de Alta Performance Reconfigurables

El objetivo de este trabajo es presentar la arquitectura y las características del sistema para el estudio de las nuevas técnicas de procesamiento paralelo en el desarrollo de sistemas de cómputo de aplicaciones específicas de altas prestaciones. Como opción en la optimización del rendimiento y la reducción en los consumos energéticos se propone una alternativa que representa un nuevo paradigma en el desarrollo de supercomputadoras. Dicha alternativa, viene de la mano de las tecnologías de los arreglos de las composiciones lógicas formando parte del cómputo de alta rendimiento reconfigurable (HPCR).

CSP Hybrid Space Computing

Space is a hazardous environment for electronic systems; therefore, the traditional approach to space computing relies upon radiation-hardened electronics, which are characteristically more expensive, larger, less energy-efficient, and generations behind in performance and functionality of modern commercial processors. Conversely, modern commercial processors, while providing the utmost in performance and energy efficiency, are susceptible to space radiation. The desire for more autonomous missions combined with growing demands for more detailed products from advanced sensors have challenged organizations to stay relevant by “doing more with less” to meet future requirements. To meet this need, researchers at the National Science Foundation Center for High-Performance Reconfigurable Computing have developed a new design concept in hybrid space computing that features fixed and reconfigurable processor architectures merged with an innovative system design that is a combination of three technologies: commercial devices, radiation-hardened components, and fault-tolerant computing. This concept has culminated in the development of several novel research platforms, most significantly the CSPv1 flight computer, which was successfully deployed on the International Space Station. Before flight, this CSPv1 design was analyzed and tested on the ground with reliability models and radiation tests. Recent in-flight data have validated the design and shown that CSPv1 exceeds reliability expectations predicted by models.

RECONFIGURABLE FPGA BASED SOFT-CORE PROCESSOR FOR SIMD APPLICATIONS

Objective: The prospective need of SIMD (Single Instruction and Multiple Data) applications like video and image processing in single system requires greater flexibility in computation to deliver high quality real time data. This paper performs an analysis of FPGA (Field Programmable Gate Array) based high performance Reconfigurable OpenRISC1200 (ROR) soft-core processor for SIMD.Methods: The ROR1200 ensures performance improvement by data level parallelism executing SIMD instruction simultaneously in HPRC (High Performance Reconfigurable Computing) at reduced resource utilization through RRF (Reconfigurable Register File) with multiple core functionalities. This work aims at analyzing the functionality of the reconfigurable architecture, by illustrating the implementation of two different image processing operations such as image convolution and image quality improvement. The MAC (Multiply-Accumulate) unit of ROR1200 used to perform image convolution and execution unit with HPRC is used for image quality improvement.Result: With parallel execution in multi-core, the proposed processor improves image quality by doubling the frame rate up-to 60 fps (frames per second) with peak power consumption of 400mWatt. Thus the processor gives a significant computational cost of 12ms with a refresh rate of 60Hz and 1.29ns of MAC critical path delay.Conclusion:This FPGA based processor becomes a feasible solution for portable embedded SIMD based applications which need high performance at reduced power consumptions

A scalable CNN architecture and its application to short exposure stellar images processing on a HPRC

A CNN-based algorithm for short exposure image processing and an application-specific computing architecture developed to accelerate its execution are presented. Algorithm is based on a flexible and scalable Cellular Neural Networks (CNN) architecture specifically designed to optimize the projection of CNN kernels on a programmable circuit. The objective of the proposed algorithm is to minimize the adverse effect that atmospheric disturbance has on the images obtained by terrestrial telescopes. Algorithm main features are that it can be adapted to the detection of several astronomical objects and it supports multi-stellar images. The implementation platform made use of a High Performance Reconfigurable Computer (HPRC) combining general purpose standard microprocessors with custom hardware accelerators based on FPGAs, to speed up execution time. The hardware/software partitioning and co-design process have been carried out using high level design tools, instead of traditional Hardware Description Languages (HDLs). Results are presented in terms of circuit area/speed, processing performance and output quality.

FPGA-based High-Performance Reconfigurable Computer

FPGA-based High-Performance Reconfigurable Computer

A Communication-aware Scheduling Algorithm for Hardware Task Scheduling Model on FPGA-based Reconfigurable Systems

Task Scheduling is an important aspect of high performance reconfigurable computing. Most of the heuristics of this NP-hard problem are based on a simple abstract model of FPGA and have little investigation into optimizing data communication which influences the system performance importantly. To solve this problem, a Communication-aware Maximum Adjacent Edges (CA-MAE) algorithm based on 2D reconfigurable model is proposed, which could reduce communication distance during scheduling and enhance the system performance. The experimental results show that CA-MAE reduces communication cost by 17%.

Optimization Of Matching and Scheduling On Heterogeneous CPU/FPGA Architectures

The continuous improvements in several application domains such as telecommunications, aerospace or multimedia lead to additional design constraints on power budget and architecture scalability. The heterogeneous CPU/FPGA (Central Processing Unit/Field-Programmable Gate Array) architecture is one of the most promising solutions in this context leading to high performance reconfigurable computing. In such systems, multi-core processors (CPU) provide high computation rates while the reconfigurable logic (FPGA) offers high performance and adaptability to the application real-time constraints. However, there is a lack of CAD (Computer Aided Design) tools able to deal with the development of applications on such heterogeneous systems. This research investigates the problem of the optimisation of static and run-time task mapping on a real-time computing system CPU/FPGA used to implement intimately coupled hardware and software models.

Efficient Mapping of Task Graphs onto Reconfigurable Hardware Using Architectural Variants

High-performance reconfigurable computing involves acceleration of significant portions of an application using reconfigurable hardware. Mapping application task graphs onto reconfigurable hardware is, therefore, of rising attention. In this work, we approach the mapping problem by incorporating multiple architectural variants for each hardware task; the variants reflect tradeoffs between the logic resources consumed and the task execution throughput. We propose a mapping approach based on the genetic algorithm, and show its effectiveness for random task graphs as well as an N-body simulation application, demonstrating improvements of up to 78.6 percent in the execution time compared with choosing a fixed implementation variant for all tasks. We then validate our methodology through experiments on real hardware, an SRC-6 reconfigurable computer.

Optimization of Shared High-Performance Reconfigurable Computing Resources

In the field of high-performance computing, systems harboring reconfigurable devices, such as field-programmable gate arrays (FPGAs), are gaining more widespread interest. Such systems range from supercomputers with tightly coupled reconfigurable hardware to clusters with reconfigurable devices at each node. The use of these architectures for scientific computing provides an alternative for computationally demanding problems and has advantages in metrics, such as operating cost/performance and power/performance. However, performance optimization of these systems can be challenging even with knowledge of the system’s characteristics. Our analytic performance model includes parameters representing the reconfigurable hardware, application load imbalance across the nodes, background user load, basic message-passing communication, and processor heterogeneity. In this article, we provide an overview of the analytical model and demonstrate its application for optimization and scheduling of high-performance reconfigurable computing (HPRC) resources. We examine cost functions for minimum runtime and other optimization problems commonly found in shared computing resources. Finally, we discuss additional scheduling issues and other potential applications of the model.

PERFORMANCE EVALUATION OF OR1200 PROCESSOR WITH EVOLUTIONARY PARALLEL HPRC USING GEP

In this fast computing era, most of the embedded system requires more computing power to complete the complex function/ task at the lesser amount of time. One way to achieve this is by boosting up the processor performance which allows processor core to run faster. This paper presents a novel technique of increasing the performance by parallel HPRC (High Performance Reconfigurable Computing) in the CPU/DSP (Digital Signal Processor) unit of OR1200 (Open Reduced Instruction Set Computer (RISC) 1200) using Gene Expression Programming (GEP) an evolutionary programming model. OR1200 is a soft-core RISC processor of the Intellectual Property cores that can efficiently run any modern operating system. In the manufacturing process of OR1200 a parallel HPRC is placed internally in the Integer Execution Pipeline unit of the CPU/DSP core to increase the performance. The GEP Parallel HPRC is activated /deactivated by triggering the signals i) HPRC_Gene_Start ii) HPRC_Gene_End. A Verilog HDL(Hardware Description language) functional code for Gene Expression Programming parallel HPRC is developed and synthesised using XILINX ISE in the former part of the work and a CoreMark processor core benchmark is used to test the performance of the OR1200 soft core in the later part of the work. The result of the implementation ensures the overall speed-up increased to 20.59% by GEP based parallel HPRC in the execution unit of OR1200.

Virtualization of reconfigurable coprocessors in HPRC systems with multicore architecture

HPRC (High-Performance Reconfigurable Computing) systems include multicore processors and reconfigurable devices acting as custom coprocessors. Due to economic constraints, the number of reconfigurable devices is usually smaller than the number of processor cores, thus preventing that a 1:1 mapping between cores and coprocessors could be achieved. This paper presents a solution to this problem, based on the virtualization of reconfigurable coprocessors. A Virtual Coprocessor Monitor (VCM) has been devised for the XtremeData XD2000i In-Socket Accelerator, and a thread-safe API is available for user applications to communicate with the VCM. Two reference applications, an IDEA cipher and an Euler CFD solver, have been implemented in order to validate the proposed architecture and execution model. Results show that the benefits arising from coprocessor virtualization outperform its overhead, specially when code has a significant software weight.

An Evaluation of an Integrated On-Chip/Off-Chip Network for High-Performance Reconfigurable Computing

As the number of cores per discrete integrated circuit (IC) device grows, the importance of the network on chip (NoC) increases. However, the body of research in this area has focused on discrete IC devices alone which may or may not serve the high-performance computing community which needs to assemble many of these devices into very large scale, parallel computing machines. This paper describes an integrated on-chip/off-chip network that has been implemented on an all-FPGA computing cluster. The system supports MPI-style point-to-point messages, collectives, and other novel communication. Results include the resource utilization and performance (in latency and bandwidth).

An End-to-End Tool Flow for FPGA-Accelerated Scientific Computing

Editor's note:As part of their ongoing work with the National Science Foundation (NSF) Center for High-Performance Reconfigurable Computing (CHREC), the authors are developing a complete tool chain for FPGA-based acceleration of scientific computing, from early-stage assessment of applications down to rapid routing. This article provides an overview of this tool chain.—George A. Constantinides (Imperial College London) and Nicola Nicolici (McMaster University)

High Performance Reconfigurable Computer Systems on the base of FPGA technology

AbstractThe paper covers the problems of design and creation of high performance reconfigurable computer systems based on FPGA technology. Authors compare performance values of up-to-date cluster systems and original reconfigurable computer systems, suggest principles of design of modular scalable reconfigurable computer systems with high real performance and give examples of various implementations of such systems. In addition, principles of applications development are suggested. Also the paper covers the problems of reconfigurable system programming. There is viewed an original complex of system software for reconfigurable system programming.

Molecular Dynamics Simulations on High-Performance Reconfigurable Computing Systems

The acceleration of molecular dynamics (MD) simulations using high-performance reconfigurable computing (HPRC) has been much studied. Given the intense competition from multicore and GPUs, there is now a question whether MD on HPRC can be competitive. We concentrate here on the MD kernel computation: determining the short-range force between particle pairs. In one part of the study, we systematically explore the design space of the force pipeline with respect to arithmetic algorithm, arithmetic mode, precision, and various other optimizations. We examine simplifications and find that some have little effect on simulation quality. In the other part, we present the first FPGA study of the filtering of particle pairs with nearly zero mutual force, a standard optimization in MD codes. There are several innovations, including a novel partitioning of the particle space, and new methods for filtering and mapping work onto the pipelines. As a consequence, highly efficient filtering can be implemented with only a small fraction of the FPGA's resources. Overall, we find that, for an Altera Stratix-III EP3ES260, 8 force pipelines running at nearly 200 MHz can fit on the FPGA, and that they can perform at 95% efficiency. This results in an 80-fold per core speed-up for the short-range force, which is likely to make FPGAs highly competitive for MD.