Kiyoshi FURUYA Edward J. McCLUSKEY
A method to analyze two-pattern test capabilities of autonomous test pattern generator (TPG) circuits for use in built-in self-testing are described. The TPG circuits considered here include arbitrary autonomous linear sequential circuits in which outputs are directly fed out from delay elements. Based on the transition matrix of a circuit, it is shown that the number of distinct transitions in a subspace of state variables can be obtained from rank of the submatrix. The two-pattern test capabilities of LFSRs, cellular automata, and their fast parallel implementation are investigated using the transition coverage as a metric. The relationships with dual circuits and reciprocal circuits are also mentioned.
Saneaki TAMAKI Michitaka KAMEYAMA
Design of high-speed digital circuits such as adders and multipliers is one of the most important issues to implement high performance VLSI systems. This paper proposes a new multiple-valued code assignment algorithm to implement locally computable combinational circuits for k-ary operations. By the decomposition of a given k-ary operation into unary operations, a code assignment algorithm for k-ary operations is developed. Partition theory usually used in the design of sequential circuits is effectively employed for optimal code assignment. Some examples are shown to demonstrate the usefulness of the proposed algorithm.
Harufusa KONDOH Hiromi NOTANI Hideaki YAMANAKA Keiichi HIGASHITANI Hirotaka SAITO Isamu HAYASHI Yoshio MATSUDA Kazuyoshi OSHIMA Masao NAKAYA
A new shared multibuffer architecture for high-speed ATM (Asynchronous Transfer Mode) switch LSIs is described. Multiple buffer memories are located between two crosspoint switches. By controlling the input-side crosspoint switch so as to equalize the utilization rate of each buffer memory, these multiple buffer memories can be recognized as a single large shared buffer memory. High utilization efficiency of buffer memory can thus be achieved, and the cell loss ratio is minimized. By accessing the buffer memories in parallel via crosspoint switches, the time required to access the buffer memories is greatly reduced. This feature enables high-speed operation of the switch. The shared multibuffer architecture was implemented in a switch LSI using 0.8-µm BiCMOS process technology. Experimental results revealed that this chip can operate at more than 125 MHz. Bit-sliced eight switch LSIs operating at 78 MHz construct a 622-Mb/s 88 ATM switching system with a buffer size of 1,024 ATM cells. Power consumption of the switch LSI was 3 W.
Masami NAKAJIMA Michitaka KAMEYAMA
To realize next-generation high performance ULSI processors, it is a very important issue to reduce the critical delay path which is determined by a cascade chain of basic gates. To design highly parallel digital operation circuits such as an adder and a multiplier, it is difficult to find the optimal code assignment in the non-linear digital system. On the other hand, the use of the linear concept in the digital system seems to be very attractive because analytical methods can be utilized. To meet the requirement, we propose a new design method of highly parallel linear digital circuits for unary operations using the concept of a cycle and a tree. In the linear digital circuit design, the analytical method can be developed using a representation matrix, so that the search procedure for optimal locally computable circuits becomes very simple. The evaluations demonstrate the usefulness of the circuit design algorithm.
We propose a method of diagnosing any logical fault in combinational circuits through a repetition of the single fault-net location procedure with the aid of probing, called SIFLAP-G. The basic idea of the method has been obtained through an observation that a single error generated on a fault-net often propagates to primary outputs under an individual test even though multiple fault-nets exist in the circuit under test. Therefore, candidates for each fault-net are first deduced by the erroneous path tracing under the single fault-net assumption and then the fault-net is found out of those candidates by probing. Probing internal nets is done only for some of the candidates, so that it is possible to greatly decrease the number of nets to be probed. Experimental results show that the number seems nearly proportional to the number of fault-nets (about 35 internal nets per fault-net), but almost independent of the type of faults and the circuit size.
Seungjik LEE Jaeho SHIN Seiichi NOGUCHI
In this letter, we study on the sensitivity to the electrical stimulus pulse for biomedical electronics for the purpose to make a tactile vision substitution system for binds. We derive the equivalent circuit of finger by measuring sensitive voltages with various touch condition and various DC voltage. And we consider to the sensitivity of finger against electrical stimulus pulse. In order to convert the sense of sight to tactile sense, we consider four types of touch condition and various types of pulse. It is shown that the sensitivity of finger to electrical stimulus pulse is determined by duty-ratio, frequency, hight of pulse and the type of touch condition. In the case that duty-ratio is about 20%, frequency is within about 60-300Hz and touch condition is A-4 type, the sensitive voltage becomes the lowest. With this result, a tactile vision substitution system can be developed and the system will be used to transfer various infomations to blinds without paper.
The authors examined the effect of feedbacking information on learners of their test results obtained through computerized tests. The learner's acceptability of computerized tests was revealed to be improved by distribution and explanation of newly devised feedback charts including data on one's response history and response latency during computerized testing that was carried out in formative evaluation. The feedback chart composed of graphic representation of relationship between degree of difficulty of each question and its response latency got a particularly high evaluation among learners. It was revealed that types of feedback chart that stood highest in learner's estimate varied with the learner traits. This observation will serve to develop educational systems that incorporate computerized tests into school lessons.
Keiji KONISHI Yoshiaki SHIRAO Hiroaki KAWABATA Toshikuni NAGAHARA Yoshio INAGAKI
A laser system which has a mirror outside of it to feedback a delayed output has been described by the Maxwell-Bloch equations with time delay. It is shown that a chaotic behavior in the equations can be controlled by using a OPF control algorithm. Our numerical simulation indicates that the chaotic behavior is stabilized on 1, 2 periodic unstable orbits.
Yuzo TAKAMATSU Taijiro OGAWA Hiroshi TAKAHASHI
In our recent work, a forward test generation method for sequential circuits by using a single time frame was proposed. In order to improve the effectiveness of the method, we introduced an extended mode which can handle the two time frames for a hard-to-test fault and a state escaping phase which can detect a sequence of unsuitable states for test generation. The experimental results show that the improved method is effective in generating higher coverage tests with a small number of tests.
Masahiro HASHIMOTO Eiji FUJIWARA
Since semiconductor memory chip has been growing rapidly in its capacity, memory testing has become a crucial problem in RAMs. This paper proposes a new RAM test algorithm, called generalized marching test (GMT), which detects static and dynamic pattern sensitive faults (PSF) in RAM chips. The memory array with N cells is partitioned into B sets in which every two cells has a cell-distance of at least d. The proposed GMT performs the ordinary marching test in each set and finally detects PSF having cell-distance d. By changing the number of partitions B, the GMT includes the ordinary marching test for B1 and the walking test for BN. This paper demonstrates the practical GMT with B2, capable of detecting PSF, as well as other faults, such as cell stuck-at faults, coupling faults, and decoder faults with a short testing time.
The design of complex VLSI systems relies more and more heavily on scientific computing for numerical simulation and configuration/performance optimization. Especially, computer simulation is becoming a component of VLSI design methodology, for which a variety of computation evolutions have been accomplished for the past two decades. There are many different forms of simulation which are used for verification of VLSI design at various stages of the whole design process. They may be classified into functional or behavioral simulators, register transfer level (RTL) simulators, gate-level logic, or simply logic, simulators, timing simulator, circuit simulators, device simulator, and process simulators. Among these simulation tasks, a series of logic, timing, and circuit simulation is most strongly related to the design stage which deals with logic/electric waveform performance of VLSI circuits. This article surveys the state of the art of VLSI simulation, putting stress mainly on the domain of logic, timing, and circuit simulation, since the reader of the Transactions may be interested exclusively in this field.
Yoshihiro FUJITA Nobuyuki YAMASHITA Shin'ichiro OKAZAKI
This paper describes the architecture and simulated performance of a proposed Integrated Memory Array Processor (IMAP). The IMAP is an LSI which integrates a large capacity memory and a one dimensional SIMD processor array on a single chip. The IMAP holds, in its on-chip memory, data which at the same time can be processed using a one dimensional SIMD processor integrated on the same chip. All processors can access their individual parts of memory columns at the same time. Thus, it has very high processor-memory data transfer bandwidth, and has no memory access bottleneck. Data stored in the memory can be accessed from outside of the IMAP via a conventional memory interface same as a VRAM. Since the SIMD processors on the IMAP are configured in a one dimensional array, multiple IMAPs could easily be connected in series to create a larger processor and memory configuration. To estimate the performance of such an IMAP, a system architecture and instruction set were first defined, and on the basis of those two, a simulator and an assembly language were then developed. In this paper, simulation results are presented which indicate the performance of an IMAP in both image processing and artificial neural network calculations.
We observed a ship as a radar target embedded in sea clutter using a millimeter wave radar. The shape of the ship and sea clutter were discriminated by using texture analysis in image processing. As a discriminator, a nonlinear transformation of a local pattern was defined to deal with high order statistics.
Takao WATANABE Masakazu AOKI Katsutaka KIMURA Takeshi SAKATA Kiyoo ITOH
The advantages of a neuro-chip architecture based on a DRAM are demonstrated through a discussion of the general issuse regarding a memory based neuro-chip architecture and a comparison with a chip based on an SRAM. The performance of both chips is compared assuming digital operation, a 1.5-V supply voltage, a 106-synapse neural network capability, and a 0.5-µm CMOS design rule. The use of a one-transistor DRAM cell array for the storage of synapse weights results in a chip 55% smaller than an SRAM based chip with the same 8-Mbit memory capacity and the same number of processing elements. No additional operations for refreshing the DRAM cell array are necessary during the processing of the neural networks. This is because all the synapse weights in the array are transferred to the processing elements during the processing and the DRAM cells in the array are automatically refreshed when they are selected. The precharge operation of the DRAM cell array degrades the processing speed, however a processing speed of 1.37 GCPS is expected for the DRAM based chip. That speed is comparable to the 1.71 GCPS for the SRAM based chip with the same 256 parallel-processing elements. A DRAM cell array has the additional advantage of lower power dissipation in this specific usage for the neuro-chip. The dynamic operation of the DRAM cell array results in a 10% lower operating power dissipation than a chip using an SRAM cell array at the same processing speed of 1.37 GCPS. That lower operating power dissipation enables a DRAM based chip to run on a 1.5-V dry cell for longer under intermittent daily use even though the SRAM cell array has little power dissipation in data-holding mode.
Masayuki OKUNO Akio SUGITA Tohru MATSUNAGA Masao KAWACHI Yasuji OHMORI Katsumi KATOH
A strictly nonblocking 88 matrix switch was designed and fabricated using silica-based planar lightwave circuits (PLC) on a silicon substrate. The average insertion loss was 11 dB in the TE mode and 11.3 dB in the TM mode. The average switch element extinction ratio was 16.7 dB in the TE mode and 17.7 dB in the TM mode. The accumulated crosstalk was estimated to be 7.4 dB in the TE mode and 7.6 dB in the TM mode. The driving power of the phase shifter required for switching was about 0.5 W and the polarization dependence of the switching power was 4%. The switching response time was 1.3 msec. The wavelength range with a switch extinction ratio of over 15 dB was 1.31 µm30 nm.
Takayuki MORISHITA Youichi TAMURA Takami SATONAKA Atsuo INOUE Shin-ichi KATSU Tatsuo OTSUKI
We have developed a digital coprocessor with a dynamically reconfigurable pipeline architecture specified for a layered neural network which executes on-chip learning. The coprocessor attains a learning speed of 18 MCUPS that is approximately twenty times that of the conventional DSP. This coprocessor obtains expansibility in the calculation through a larger multi-layer, network by means of a network decomposition and a distributed processing approach.
Scaling-down of MOSFETs (metal-oxide-semiconductor field effect transistors can be divided to semi-classical and quantum mechanical one. In the regime of semi-classical scaling-down the behavior of electrons and holes can be well described with the effective mass approximation and in the regime of quantum mechanical scaling-down the characteristics of electrons and holes as wave becomes markedly. The minimum size limit of MOSFETs scaled down in semi-classical regime is mainly determined by the subthreshold characteristics and the short channel effect on the threshold voltage and 0.1 µm will be the minimum channel length from practical viewpoints. Scaling down of MOSFETs enhances their operational speed, but the substrates with high resistivity which are often used in SOI (silicon on insulator) substrates result longer dielectric relaxation time. While the dielectric relaxation time becomes longer than the reciprocal of signal frequency, the semiconductors work as lossy dielectrics and may lead to new types of dynamic circuits. Modification of material properties utilizing the wave nature of electrons is an illustration of quantum mechanical way to improve characteristics of MOSFETs. Suppression of optical phonon scattering of two dimensional electrons by introducing two dimensional array of quantum dots into substrates is expected to improve high field characteristics of material. Brillouin zone folding is another way to control the band structure of materials, especially to make the indirect transition band structure to the direct transition band structure. Heat transfer from a chip severely limits the number of devices which can be integrated on the chip. Reduction of signal charge to electronic elementary charge, that is quantum limit, is expected to be useful for realization of nano-power electronics.
Intermittent chaos was observed in the silicon thyristor circuit without external elements of L and C, under the condition of ac excitation at the anode. Lorenz plot reconstructed from the experimental waveform and the numerical simulation of this kind of intermittency fairly agreed with each other.
Current status and critical issues of the material and device technology towards constructing new architecture LSIs based on quantum-mechanical principles are reviewed in an attempt to draw attention of systems workers to the field. Limitations of the present-day LSI architecture are discussed from the viewpoints of material science and device physics. New quantum mechanical phenomena in the quantum structures are reviewed. Then, key material and processing issues for fabrication of desired quantum structures are briefly discussed. Finally, the basic operation principles the quantum devices and possible architectures of quantum LSIs are discussed.
Toshiaki TAKEDA Hiroki MIZOE Koichiro KISHI Takahide MATSUOKA
To investigate necessary conditions for the object recognition by simulations using neural network models is one of ways to acquire suggestions for understanding the neuronal representation of objects in the brain. In the present study, we trained a three layered neural network to form a geometrical feature representation in its output layer using back-propagation algorithm. After training using 73 learning examples, 65 testing patterns made by various combinations of above features could be recognized with the network at a rate of 95.3% appropriate response. We could classify four types of hidden layer units on the basis of effects on the output layer.