Hardware evolution system AdAM (original) (raw)
1999, Communications of the ACM
https://doi.org/10.1145/299157.299870
Sign up for access to the world's latest research
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Abstract
AI
The paper introduces the Hardware evolution system AdAM, designed to facilitate and streamline the continuous evolution of hardware systems. AdAM leverages advanced algorithms to optimize hardware design processes, allowing for enhanced performance and adaptability in complex environments. The impact of AdAM on the efficiency of hardware evolution is quantitatively assessed, showing significant improvements over traditional methods.
Related papers
Adaptive Interface Based on FPGA implemented Artificial Neural Network
The goal of this work is to build smart interfaces with learning and adaptive capability. The key element of the learning and adaptive behavior are artificial neural network (ANIN) blocks, implemented in FPGA using the System Generator tool for Simulink developed by Xilinx Inc. This tool allow the easy generation of hardware Description Language (I1DL) code from a system representation in Simulink. This VHDL design can then bc synthcsized for implementation in the Xilinx family of FPGA devices. The off-chip learning task is performed using Matlab and the ANN's wcights are transferred automatically from Matlab workspace to weights memory.
Dynamically Reconfigurable Hardware for Evolving Bio-Inspired Architectures
Emerging Trends and Applications
During the last few years, reconfigurable computing devices have experienced an impressive development in their resource availability, speed, and configurability. Currently, commercial FPGAs offer the possibility of self-reconfiguring by partially modifying their configuration bit-string, providing high architectural flexibility, while guaranteeing high performance. On the other hand, we have bio-inspired hardware, a large research field taking inspiration from living beings in order to design hardware systems, which includes diverse approaches like evolvable hardware, neural hardware, and fuzzy hardware. Living beings are well known for their high adaptability to environmental changes, featuring very flexible adaptations at several levels. Bio-inspired hardware systems require such flexibility to be provided by the hardware platform on which the system is implemented. Even though some commercial FPGAs provide enhanced reconfigurability features such as partial and dynamic reconfigu...
Custom architectures for fuzzy and neural networks controllers
2011
Standard hardware, dedicated microcontroller or application specific circuits can implement fuzzy logic or neural network controllers. This paper presents efficient architecture approaches to develop controllers using specific circuits. A generator uses several tools that allow translating the ...
Dynamically reconfigurable bio-inspired hardware
During the last several years, reconfigurable computing devices have experienced an impressive development in their resource availability, speed, and configurability. Currently, commercial FPGAs offer the possibility of self-reconfiguring by partially modifying their configuration bitstream, providing high architectural flexibility, while guaranteeing high performance. These configurability features have received special interest from computer architects: one can find several reconfigurable coprocessor architectures for cryptographic algorithms, image processing, automotive applications, and different general purpose functions. On the other hand we have bio-inspired hardware, a large research field taking inspiration from living beings in order to design hardware systems, which includes diverse topics: evolvable hardware, neural hardware, cellular automata, and fuzzy hardware, among others. Living beings are well known for their high adaptability to environmental changes, featuring very flexible adaptations at several levels. Bio-inspired hardware systems require such flexibility to be provided by the hardware platform on which the system is implemented. In general, bio-inspired hardware has been implemented on both custom and commercial hardware platforms. These custom platforms are specifically designed for supporting bio-inspired hardware systems, typically featuring special cellular architectures and enhanced reconfigurability capabilities; an example is their partial and dynamic reconfigurability. These aspects are very well appreciated for providing the performance and the high architectural flexibility required by bio-inspired systems. However, the availability and the very high costs of such custom devices make them only accessible to a very few research groups. Even though some commercial FP-GAs provide enhanced reconfigurability features such as partial and dynamic reconfiguration, their utilization is still in its early stages and they are not well supported by FPGA vendors, thus making their use difficult to include in existing bio-inspired systems. In this thesis, I present a set of architectures, techniques, and methodologies for benefiting from the configurability advantages of current commercial FPGAs in the design of bio-inspired hardware systems. Among the presented architectures there are neural networks, spiking neuron models, fuzzy systems, cellular automata and random boolean networks. For these architectures, I propose several adaptation techniques for parametric and topological adaptation, such as hebbian learning, evolutionary and co-evolutionary algorithms, and particle swarm optimization. Finally, as case study I consider the implementation of bio-inspired hardware systems in two platforms: YaMoR (Yet another Modular Robot) and ROPES (Reconfigurable Object for Pervasive Systems); the development of both platforms having been co-supervised in the framework of this thesis.
Evolving an adaptive digital filter
Proceedings. The Second NASA/DoD Workshop on Evolvable Hardware, 2000
One important feature of signal processing is coping with noise. In a non-adaptive filter, characteristics of the filter may be refined to remove noise. One method of achieving this is to use evolution to decide the filter characteristics. However, if the noise level is sufficient or the input signal is not of the required type for the output signal required, then a satisfactory output signal may not be achievable. To be able to achieve the required output signal for a wide range of input signals and noise, it is desirable to be able to adjust both the characteristics and the type of the filter. In this way the resulting filter may be said to be an adaptive filter. In this paper we propose an on-chip solution for an adaptive digital filter using an onchip evolvable hardware method. We highlight a challenge within evolvable hardware for adaptive designs and that is to find efficient ways in which sufficient genetic material will be available to the evolution process. This problem appears when the evolution process is automatically restarted so as to adapt to a change in the environment.
Today micro-controllers and signal processors are the standard-implementing platform for embedded controllers. But often their performance can not keep pace with requirements for demanding applications, for example in mechatronic systems. Usually, in this case the microcontroller is replaced by a faster one. But this doesn't avoid software problems arising in realtime applications even by using high level languages and always by using assembler language. Further problems are concerned with higher clock frequencies and energy consumption. On the other hand FPGAs are very well suited for high speed applications, because information processing can be performed in parallel. But they usually lack common mathematical function libraries, so that control and filter algorithms have to be implemented in a time consuming and expensive process. Today's FPGAs offer a huge number of gates, so that even a processor core can be implemented inside. This enables a combination of hardware and software signal processing within only one IC. Thereby the available software can be used further on and only some algorithms requiring high performance have to be implemented in hardware. Furthermore some aspects of reconfigurable computing are treated.
A dedicated very low power analog VLSI architecture for smart adaptive systems
Applied Soft Computing, 2004
This paper deals with analog VLSI architectures addressed to the implementation of smart adaptive systems on silicon. In particular, we addressed the implementation of artificial neural networks with on-chip learning algorithms with the goal of efficiency in terms of scalability, modularity, computational density, real time operation and power consumption. We present the analog circuit architecture of a feed-forward network with on-chip weight perturbation learning in CMOS technology. Novelty of the approach lies in the circuit implementation of the feed-forward neural primitives and on the overall analog circuit architecture. The proposed circuits feature very low power consumption and robustness with respect to noise effects. We extensively tested the analog architecture with simulations at transistor level by using the netlist extracted from the physical design. The results compare favourably with those reported in the open literature. In particular, the architecture exhibits very high power efficiency and computational density and remarkable modularity and scalability features. The proposed approach is aimed to the implementation of embedded intelligent systems for ubiquitous computing. artificial vision, pattern recognition, etc. We refer to the "soft computing" techniques like artificial neural networks, cellular networks, fuzzy logic, etc.;
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.
References (3)
- Hikage, T., Hemmi, H., and Shimohara, K. Hardware evolution system introducing dominant and recessive heredity. In Proceedings of ICES '96 (Tsukuba, Japan, Oct. 7-8). LNCS 1259, Springer, 1996, 423-436.
- Hikage, T., Hemmi, H., and Shimohara, K. Progressive evolution model using a hardware evolution system. In Proceeding ofAROB '97 (Oita, Japan, Feb. 18-20), 1997, 18-21.
- Mizoguchi, J., Hemmi, H., and Shimohara, K. Production genetic algo- rithms for automated liardware design through an evolutionary process. In Proceeding of ICEC '94 (Orlando, Fla. June 27-29), 1994, 66l-<564.
Related papers
Implementation of adaptive logic networks on an FPGA board
Configurable Computing: Technology and Applications, 1998
This work is part of a project that studies the implementation of neural network algorithms in reconfigurable hardware as a way to obtain a high performance neural processor. The results for Adaptive Logic Network (ALN) type binary networks with and without learning in hardware are presented.
Reconfigurable architectures for system-level applications of adaptive computing.” ISI
2013
Adaptive computing systems use FPGAs for custom hardware acceleration in high performance and real-time applications. Unlike single purpose dedicated hardware approaches, the reusable nature of the technology introduces system design tradeoffs that must balance processing density, memory, and I/O bandwidth, not to mention more subtle issues such as ease of pro gramming, debugging, and physical integration into real-world systems. This paper describes results from the DARPA-funded SLAAC project, which developed three generations of adaptive computing systems for a diverse set of challenging signal processing applications.
Implementation of adaptive logic networks on an FPGA board
1998
This work is part of a project that studies the implementation of neural network algorithms in reconfigurable hardware as a way to obtain a high performance neural processor. The results for Adaptive Logic Network (ALN) type binary networks with and without learning in hardware are presented.
Reconfigurable Architectures for System Level Applications of Adaptive Computing
VLSI Design, 2000
The System Level Applications of Adaptive Computing (SLAAC) project is defining an open, distributed, scalable, adaptive computing systems architecture based on a highspeed network cluster of heterogeneous, FPGA-accelerated nodes. Two reference implementations of this architecture are being created. The Research Reference Platform (RRP) is a Myrinet TM cluster of PCs with PCI-based FPGA accelerators (SLAAC-1). The Deployable Reference Platform (DRP) is a Myrinet cluster of PowerPC TM nodes with VME-based FPGA accelerators (SLAAC-2) and a commercial 6U-VME quad-PowerPC board (CSPI M2641S) serving as the carrier. A key strategy proposed for successful ACS technology insertions is source-code compatibility between the RRP and DRP platforms. This paper focuses on the development of the SLAAC-1 and SLAAC-2 accelerators and how the network-centric SLAAC system-level architecture has shaped their designs. A preliminary mapping of a Synthetic Aperture Radar/Automatic Target Recognition (SAR/ATR) algorithm to SLAAC-2 is also discussed.
An evolvable hardware FPGA for adaptive hardware
Proceedings of the 2000 Congress on Evolutionary Computation. CEC00 (Cat. No.00TH8512)
Can we realise the opportunities that lie in design by evolution by using traditional technologies or are there better technologies which will allow us to fully realise the potential inherent in evolvable hardware? In this paper we consider the characteristics of evolvable hardware-especially for adaptive design, and discuss the demands that these characteristics place on the underlying technology. We suggest a potential alternative to today's FPGA technology. The proposed architecture is particularly focused at reducing the genotype required for a given design by reducing the configuration data required for unused routing resources and allowing partial configuration down to a single CLB. In addition, to support adaptive hardware, self-reconfiguration is enabled.
A fast Reconfigurable 2D HW core architecture on FPGAs for evolvable Self-Adaptive Systems
2011 NASA/ESA Conference on Adaptive Hardware and Systems (AHS), 2011
Modern FPGAs with Dynamic and Partial Reconfiguration (DPR) feature allow the implementation of complex, yet flexible, hardware systems. Combining this flexibility with evolvable hardware techniques, real adaptive systems, able to reconfigure themselves according to environmental changes, can be envisaged. In this paper, a highly regular and modular architecture combined with a fast reconfiguration mechanism is proposed, allowing the introduction of dynamic and partial reconfiguration in the evolvable hardware loop. Results and use case show that, following this approach, evolvable processing IP Cores can be built, providing intensive data processing capabilities, improving data and delay overheads with respect to previous proposals. Results also show that, in the worst case (maximum mutation rate), average reconfiguration time is 5 times lower than evaluation time.
Autonomous System on a Chip Adaptation through Partial Runtime Reconfiguration
This paper presents a proto-type autonomous signal processing system on a chip. The system is architected such that high performance digital signal processing occurs in the FPGA's configurable logic, while resulting higher level data products are monitored by cognitive algorithms residing on an embedded processor. The cognitive algorithms develop situational awareness about the platform's environment, and use this information to modify and tune signal processing in real-time using active partial reconfiguration. This system was realized on a Xilinx Virtex4 FX 100 device on a pulse parameter measurement application utilizing a Bayesian Network cognitive algorithm. Changes in the RF environment were correctly detected 96.7% of the time and mitigation filters which resulted in at least 3dB interference rejection improvement were instanced 81% of the time. This system realizes a 71.4x reduction in size compared to static implementations and a 26-43x reduction in reaction times c...
COMPUTING SYSTEMS BASED ON ADAPTIVE RECONFIGURABLE ARCHITECTURES
2012
ABSTRACT Imagine further a computing system that performs better according to a user's preferred goal the longer it runs an application. Such an architecture will enable, for example, a hand-held radio or a cell phone that can run cooler the longer the connection time. Moreover, Systems on a Chip (SoC) can draw various benefits, such as adaptability and efficient acceleration of compute-intensive tasks from the inclusion of reconfigurable hardware as a system component.
On the properties of artificial development and its use in evolvable hardware
2009 IEEE Symposium on Artificial Life, 2009
The design of a new biologically inspired artificial developmental system is described in this paper. In general, developmental systems converge slower and are more computationally expensive than direct evolution. However, the performance trends of development indicate that the full benefit of development will arise with larger and more complex problems that exhibit some sort of regularity in their structure: thus, the aim is to evolve larger electronic systems through the modularity allowed by development. The hope is that the proposed artificial developmental system will exhibit adaptivity and fault tolerance in the future. The cell signalling and the system of Gene Regulatory Networks present in biological organisms are modelled in our developmental system, and tailored for tackling real world problems on electronic hardware. For the first time, a Gene Regulatory Network system is successfully shown to develop the complete circuit structure of a desired digital circuit without the help of another mechanism or any problem specific structuring. Experiments are presented that show the modular behaviour of the developmental system, as well as its ability to solve non-modular circuit problems.
Adaptive processor architecture - invited paper
2012 International Conference on Embedded Computer Systems (SAMOS), 2012
This paper introduces a novel methodology to adapt the microarchitecture of a processor at run-time. The goal is to tailor the internal architecture to the requirements of an application and the data to be processed. The latter parameter is normally not known at design time. This leads to the development of more general purpose processors which are capable to handle the data to be processed in any case. With the novel approach which keeps the microarchitecture of a processor flexible, the processor can start as a general purpose device and end up with a specific parameterization, comparable with application specific processor architectures. Furthermore, the increased degree offreedom which is enabled through the approach for a novel quality of processors is described.