circuits implementing the concept of Selective Fault Tolerance according to [1] are fault-tolerant for a specified subset of inputs. In this paper, a new heuristic is presented to make the method of Selective Fault To...
详细信息
ISBN:
(纸本)9781424497560
circuits implementing the concept of Selective Fault Tolerance according to [1] are fault-tolerant for a specified subset of inputs. In this paper, a new heuristic is presented to make the method of Selective Fault Tolerance applicable to industrial designs. The heuristic can be efficiently implemented by use of conventional design tools. Compared to TMR, the method, in combination with the heuristic, saves a huge amount of area redundancy and fault tolerance is adapted to the real requirements of a system specification. This is demonstrated by experimental results obtained from circuit descriptions in Verilog and a synthesis with the tool Synopsys.
Neural networks with quantized activation functions cannot adapt the quantization at the input of their first layer. Preprocessing is therefore required to adapt the range of input data to the quantization range. Such...
详细信息
ISBN:
(纸本)9798350359343
Neural networks with quantized activation functions cannot adapt the quantization at the input of their first layer. Preprocessing is therefore required to adapt the range of input data to the quantization range. Such preprocessing usually includes an activation-wise linear transformation and is steered by the properties of the training set. We suggest to include the linear transform into the training process. Using the Jet Stream Classification task and an evaluation architecture of three quantized dense layers, we document that it improves accuracy, requires the same resources as standard preprocessing, plays a role in network pruning, and is reasonably stable with respect to initialization.
Partial dynamic reconfiguration enables designs with increased functional density and lower power consumption, but on the other hand it increases the complexity of the design process. This paper describes a methodolog...
详细信息
ISBN:
(纸本)1424401844
Partial dynamic reconfiguration enables designs with increased functional density and lower power consumption, but on the other hand it increases the complexity of the design process. This paper describes a methodology and design flow for designs with dynamic reconfiguration in the DSP and control domain. The described design flow starts with a description in Matlab / Simulink that is converted to Handel-C and then compiled through VHDL to EDIF, and finally to FPGA configuration. The methodology and design flow are demonstrated on implementation examples with simple floating-point IP cores targeting the Atmel AT94K FPSLIC device.
The nonlinear binary codes have a great disadvantage of difficult enumeration of the results due to the necessity of solving a nonlinear set of equations. In the common case, the computational effort will be equal to ...
详细信息
ISBN:
(纸本)9781538657546
The nonlinear binary codes have a great disadvantage of difficult enumeration of the results due to the necessity of solving a nonlinear set of equations. In the common case, the computational effort will be equal to the effort of solving a set of nonlinear binary equations with a big number of variables. On the other hand, the nonlinear codes substantially overcome the best parameters of the linear ones. It is possible to design such nonlinear binary cyclic codes that have numbers of individually specified bits substantially bigger than can be obtained for linear ones. In this paper, we discuss properties of built-in pseudo-exhaustive test pattern generators and test compressors that employ the nonlinear binary codes. We elaborate the previous results with new simulations. The results can be used for generator or decompressor structure optimization. We formulate several topics that have to be solved in future to enable practical application of the nonlinear codes in testing.
MDCT is the basic processing component for high quality audio compression. It is also the most computationally intensive operation in the vast majority of audio compression standards. Mostly used audio standard for au...
详细信息
ISBN:
(纸本)9781424433391
MDCT is the basic processing component for high quality audio compression. It is also the most computationally intensive operation in the vast majority of audio compression standards. Mostly used audio standard for audio compression is still MP3. This paper presents new implementations of five MDCT / IMDCT architectures with different parallelization levels for MP3. Implementation utilize UMC 90 nm CMOS technology. design was optimized for low power applications. Low power libraries and clock gating technique were used for power reduction. All IP cores are capable of computing forward and backward MDCT and this feature makes them universal in multimedia SoCs for accelerating the MP3 audio compression/decompression.
This paper presents an analysis of an influence of global parametric faults (GPF) on analogue integrated circuits (AIC) time domain (TD) response features, such as overshoot, delay time, rise time, maxima and minima, ...
详细信息
ISBN:
(纸本)9781424422760
This paper presents an analysis of an influence of global parametric faults (GPF) on analogue integrated circuits (AIC) time domain (TD) response features, such as overshoot, delay time, rise time, maxima and minima, first differential maxima and minima. The novel approach is the analysis of relations and superrelations between features which are discussed in details. The presented results should increase testability and diagnosability of the global parametric faults on a production line of analogue and mixed electroniccircuits. The recently observed AIC faults belong to parametric ones and are caused by technological process. GPF influence on aforementioned features is presented with the use of an exemplary circuit. A research presented in this paper may be used as an introduction to a Fault Driven Test (FDT) diagnosis with the use of Simulation Before Test (SBT) methods.
Motivated by the fact that no existing common mode feedback block (CMFB) can work for the systems characterized with both high output impedance and large output swing, a new CMFB topology is introduced in this paper t...
详细信息
ISBN:
(纸本)9781424422760
Motivated by the fact that no existing common mode feedback block (CMFB) can work for the systems characterized with both high output impedance and large output swing, a new CMFB topology is introduced in this paper to accomplish this task. Various types of CMFB are studied, analyzed and simulated. Compared with other CMFBs, the proposed CMFB obtains the widest linear input range, smallest output common-mode level error, largest (full) output swing and has no interference to the system output impedance. All these properties entitle the presented CMFB circuit to suit all types of fully differential systems, especially for systems which require both high output impedance and large signal output swing.
Accurate analysis of CMOS gate power dissipation shows that amount of consumed energy depends on a reason of the gate switching. Number of activated inputs and type of applied signals have an influence on dynamic powe...
详细信息
ISBN:
(纸本)9781424411610
Accurate analysis of CMOS gate power dissipation shows that amount of consumed energy depends on a reason of the gate switching. Number of activated inputs and type of applied signals have an influence on dynamic power dissipation of the gate due to dynamic reconfiguration of internal gate parasitic capacitors. Therefore, authors propose new modeling of dynamic power dissipation in static CMOS gates. Accurate modeling of dynamic power dissipation needs to take into consideration changes of all input signals. So, authors introduce new measure of digital circuit activity - gate driving way - for precise modeling of power dissipation. Based on conclusions flowing from the model analysis, authors propose method for two-level low-power circuitsdesign.
The article gives a theoretical analysis of the generalized structure of precision instrumentation amplifiers (IA) of modern interfaces for sensor and diagnostic systems of robots and unmanned aerial vehicles (UAV), w...
详细信息
ISBN:
(纸本)9781509024674
The article gives a theoretical analysis of the generalized structure of precision instrumentation amplifiers (IA) of modern interfaces for sensor and diagnostic systems of robots and unmanned aerial vehicles (UAV), which operate, among others, on exposure to radiation and low temperatures. The main disadvantages of classical IA are determined from the common positions and the possible ways of their clearing are presented. The new architecture of radiation-hardened differential difference op-amp (DDA) based on "folded" cascode is developed for the application in sensor interfaces and its simulation results are given, taking into account the effect of disturbing factors.
designs of asynchronous networks-on-chip are of growing interest because a complete asynchronous implementation can solve the synchronization problems of large networks. However, asynchronous circuits suffer from the ...
详细信息
ISBN:
(纸本)9781538604724
designs of asynchronous networks-on-chip are of growing interest because a complete asynchronous implementation can solve the synchronization problems of large networks. However, asynchronous circuits suffer from the lack of proper design flows because their functionality often relies on timing constraints, which are not extensively supported by common CAD synthesis tools. This paper proposes the design and implementation of an asynchronous router architecture suitable for a network-on-chip in the context of a Vision-System-on-Chip. The developed design flow for the synthesis of asynchronous bundled-data pipelines is based on common synthesis tools and, therefore, enables high compatibility with synchronous designs and a low barrier to entry.
暂无评论