Search

Lateral size effects on surface-roughness-limited mobility in silicon-nanowire FETs are analyzed by means of a full- quantum 3-D self-consistent simulation. A statistical analysis is carried out by considering different realizations of the potential roughness at the Si-SiO2 interfaces. Nanowires with lateral section varying from 3 times 3 to 7 times 7 nm2 are considered. Effective mobility is computed by evaluating the electron density in a reduced channel region to eliminate parasitic effects from contacts. It is found that transport in wires with the smallest section is dominated by scattering due to potential fluctuations, resulting in a larger standard deviation of the effective mobility, whereas it is dominated by transverse-mode coupling in wires with larger section, resulting in a stronger influence of surface roughness at high gate voltages.

In this paper, a dual-gate Flash EEPROM cell is proposed which allows the storage of two bits with only a slight increase in the cell size with respect to the single-bit case. Extensive simulations show that the basic functions of the Flash cell, namely reading, programming and erasing are possible with a suitable setting of the applied voltages. A simplified model based on the equivalent circuit of the cell allows for a qualitative interpretation of the obtained results, providing at the same time useful hints for the optimization of the cell parameters. Moreover, a schematic of the cell array is presented and some issues related to the array organization are briefly discussed.

In this work we investigate techniques for embedding domain-specific spatial invariances into highly-constrained neural networks. This information is used to drastically reduce the number of weights which have to be determined during the learning phase, thus allowing us to apply artificial neural networks to problems characterized by a relatively small number of available examples. As an application of the proposed methodology, we study the problem of optical inspection of machined parts. More specifically, we have characterized the performance of a network created according to this strategy, which accepts images of parts under inspection at its input and issues a flag at its output which states whether the part is defective. The results obtained so far show that the proposed methodology provides a potentially relevant approach for the quality control of industrial parts, as it offers both accuracy and short software development time, when compared with a classifier implemented using a standard approach.

We designed and built an application-specific digital architecture aimed at mining fuzzy databases. The semi custom chip set is fully functional, and a 16-unit system provides a speedup of 500 times over the algorithm (in software) running on a Sun Sparc2 workstation.

In this paper we study impact ionization in a BJT, by means of a newly develope'd technique based on a deterministic one-dimensional solution of the BTE. We evaluate the collector current multiplication factor and compare it with measurements in a thinbase transistor. Impact ionization is likely to become the major physical limiting factor toward the miniaturization of silicon bipolar transistors. In fact the collector-base breakdown voltage poses a serious limit to the maximum supply voltage and to the maximum doping level inside the collector. For this reason the collector profile itself must be carefully designed, and the ionization mechanism properly predicted by means of an efficient and accurate mlodeling tool. The limitations of the drift-diffusion approach in short devices have long been recognized. Traditional alternative methods are either based on the energy transport or hydrodynamic model or on the Monte Carlo particle simulation. While the former is reasonably fast but inadequate for describing the high-energy ionizing electrons, the latter suffers a major drawback in the huge required CPU time. In recent papers a deterministic approach for finding approximate solutions of the Boltzmann Transport Equation (BTE) has been proposed, based on a Spherical Harmonics Expansion of the distribution function [l, 21. This method gave quite good results in both the homogeneous and the one-dimensional problems, providing a sufficiently accurate description of the electron energy distribution inside the device with reasonable CPU times. The purpose of this work is to extend the procedure described in [2] to account for impact ionization. We test and validate the model by simulating a one-dimensional section of a BJT with realistic impurity profiles taken from the literature [3]. The ionization model contains only one adjustable parameter, which is calibrated by comparying the ionization coefficient with Monte Carlo homogeneous simulations and with measurements obtained in long diodes. We conclude showing how the model can approximately predict the collector current multiplication factor in the B JT under investiga.tion.

The authors describe the design and implementation of an application-specific digital architecture aimed at the solution in real time of the "K nearest neighbors" algorithm for classification problems, whose computational weight is very high. The system described here is an array-processor architecture in which the data base is split among several units that at the same time apply the same operations to different data stored in local memories. The sustained memory bandwidth of a system featuring 8 units is 2560 Mb/s and it provides a speed-up of 500 times over the software algorithm running on a SUN Sparc2 workstation. >

A 3.5 ns, 64 bit, carry-lookahead adder has been designed in full-custom domino logic and manufactured in a standard 1 /spl mu/m CMOS technology featuring two metal levels. The adder has a novel array structure which represents a variant of the architecture suggested by Brent and Kung. As opposed to the latter, however, it does not require the back propagation of the signals which is necessary for the intermediate carry bits; hence only log/sub 2/ n logic levels are employed for the generation of all the carry signals. Furthermore, the structure is highly regular and modular and can be assembled with n log/sub 2/ n identical cells with a fan-out of 2. Therefore, a compact circuit is achieved with excellent performance. The occupied area is 3370/spl times/482 /spl mu/m/sup 2/ with a worst-case 650 mW power dissipation at 100 MHz.

This paper describes the design and implementation of an application-specific digital architecture aimed at performing the mining of fuzzy database. The system described is an array-processor architecture in which the database is split among several units that at the same time apply the same operations to different data stored in local memories. A chip set implementing the architecture has been built using a semicustom approach and it is fully functional. The sustained memory bandwidth of a system featuring 8 units is 2560 Mbit/sec and it provides a speed-up of 500 times over the software algorithm running on a SUN Sparc2 workstation. >

In this paper a novel methodology for testing dedicated analog fuzzy processors is presented. It relies on small additional hardware, and features independence from the actual input-output relationship of the processor and low time for the test scan. In fact validation time does not increase exponentially with the number of inputs, as it would with an exhaustive sampling of the input-output relationship. Some simulation results show that all the most likely faults are detected.

It is the purpose of this study to verify whether or not there is a correlation between some parameters for cartilaginous injury (width, depth, and type of injury), analyzed separately, and the medium to long-term prognosis of the same. Thus, the parameters were analyzed, correlating them individually with the results obtained when a self-evaluation form was used, and the Lysholm-Gillquist score system. The small number of patients (because of the strict criteria for inclusion) does not allow for a statistically valid analysis of the results obtained. At any rate, the following considerations may be made: the depth parameter for the injury may be a valid element alone in determining a medium- to long-term prognosis; this agrees with what is reported in the literature. The parameter for width of the injury, based on our results, does not appear to be a sufficient element to determine a correlation with future prognosis. As for the type of lesion parameter, this could be correlated with the prognosis. Studies conducted in this sense with higher numbers of patients could determine the type of injury with the worst prognosis.

The authors report the results of 18 hemiarthroplasties of the shoulder performed between 1990 and 1994 using the Neer II monoblock prosthesis, emphasizing the technical and surgical problems encountered. The patients treated numbered 13 for traumatic pathology (acute fractures: 7; sequelae of fracture of the proximal epiphysis of the humerus: 6), and 5 for degenerative lesions (arthrosis: 3; rheumatoid arthritis: 2). An evaluation of the results was based on the Constant method and isokinetic testing. Results based on follow-up obtained after 2 to 6 years were satisfactory in 83% of the cases.

A total of 100 cases surgically treated for rupture of the extra-rotator cuff were evaluated by isokinetic testing. Results were evaluated after a period of time ranging from 1 to 4 years, based on the Constant system, defining the type of lesion of the cuff based on the Snyder classification system, and measuring strength postsurgery using isokinetic testing after 3 and 6 months, and 1 year. Isokinetic testing made it possible to accurately evaluate which patients are capable of obtaining better recovery, and how much time is required for physiotherapy for postoperative rehabilitation, also allowing for correction on an individual basis of any muscular imbalance.

Handwritten digit and alphabetic character recognition has been an active area of research for many years. Character recognition techniques associate a symbolic identity with the image of a character. The problem of replication of human functions by machines involves the recognition of both machine and hand-printed or cursive-written characters. Character recognition is better known as Optical Character Recognition (OCR) since it deals with recognition of optically processed characters rather than, for example, magnetically processed ones. The origin of character recognition can be found as early as in 1870, though the modern version of OCR appeared in the middle of the 1940s with the development of digital computers. Its main area of interest was data processing with application to the business world. The principal motivation for development of OCR system is the need to cope with the enormous flood of paper such as bank cheques, commercial forms, government records, mail sorting generated by the expanding society and international partnerships. Since the origin of OCR systems extensive research has been carried out and a large number of technical papers have been published by researchers in this area. Many techniques and experimental results have been published in the literature. In recent years, new classification approaches, such as neural networks, and design methodologies, such as classifier combination and parallel feature networks extraction, have renewed interests in this fields. As OCR technology seeks applications in more complicated, real time environments, the criterion for successful algorithm becomes more rigorous. New feature extraction and classification techniques are developed with parallel hardware implementation in mind. Several recognition methodologies are combined in order to improve the recognition quality. State of the art reports on character recognition research have been presented by Nagy [1968], Harmon [1972], Suen et al. [1980], Mori et al. [1984], Mantas [1986], Davis and Yall [1986], Govindan and Shivaprasad [1990], Lee and Srihari [1993].

The authors present the results they obtained in the first 100 operations performed between January 1989 and November 1990 for the treatment of lumbar sciatic pain related to disc and/or bone compression: microsurgery according to the Caspar method was used. A total of 113 spaces were submitted to surgery; discectomy at two levels was performed in 13 patients. Nerve root compression observed was based on three causes: 1) pure disc hernia (57.5%), 2) pure lateral stenosis (15.9%), 3) combined pathology (hernia+stenosis) (25.6%). In one of the cases the pathology remained unknown. At an average two-year follow-up there were 94 excellent or good results, 2 fair results, and 4 poor results. Complications included dural lacerations in 3 cases, with no sequelae. Eight patients were submitted to further surgery for recurrence of symptoms; the final results after reintervention were excellent in 7 cases, and fair in 1. The authors emphasize the advantages to using microsurgery as compared to macrosurgery, and stress a frequent finding of pure lateral stenosis or associated with disc pathology (47 out of 113 levels operated, equal to 41%), as well as the reliability of CT scan when dealing with disc pathology (97% positive diagnosis) but its unreliability when diagnosis is lateral stenosis (19 out of 47, equal to 40%).

The authors point out that when a triangular shape for the electric field in the base-collector space-charge region of an n-p-n Si BJT (bipolar junction transistor) is assumed, the electron mean energy can be calculated analytically from a simplified energy-balance equation. On this basis a nonlocal-impact-ionization model, suitable for computer-aided circuit simulation, has been obtained and used to calculate the output characteristics at constant emitter-base voltage (grounded base) of advanced devices. Provided the experimental bias-dependent value of the base parasitic resistance is accounted for in the device model, the base-collector voltage at which impact-ionization-induced snap-back occurs can be accurately predicted. >

A theoretical treatment of the current transport in MSM devices is presented. It is based upon the application of the thermionic diffusion theory to both junctions and makes it possible to relate the current density to the total voltage applied to the device by a single expression valid for any voltage above reach through. So doing, the contributions of the different transport mechanisms, i.e., thermionic injection over the energy barriers at the junctions and drift diffusion within the semiconductor, may be evaluated both for majority and minority carriers. It turns out that before the flat‐band condition the current is limited by transport phenomena in the semiconductor, while above flat band neither mechanism definitely prevails. However, in spite of the different physical models, numerical results are of the same order of magnitude as those obtained by Sze et al.

The transport properties of polycrystalline silicon films are examined and interpreted in terms of a modified grain‐boundary trapping model. The theory has been developed on the assumption of both a δ‐shaped and a uniform energy distribution of interface states. A comparison with experiments indicates that the interface states are nearly monovalent and peaked at midgap. Their density is 3.8×1012 cm−2, in accordance with carrier‐lifetime measurements performed on CVD films.

The exact-field diffuse-scattering theory of surface conduction in the inversion layer of MOS transistors has been extended to take electron gas degeneracy into account. The theory has been developed on the assumptions of spherical energy bands, and relaxation time independent of energy. The effective mobility thus obtained is compared with the corresponding results of the classical theory, showing that degeneracy is responsible for a slower decrease of mobility with increasing voltage, in the high field region. In the range of temperature between 200 and 300°K and for a carrier concentration higher than 10 12 cm −2 the theory closely agrees with the experimental low μ eff ∼ V − α T − β .

In this paper we present a generalized scaling theory which allows for an independent scaling of the FET physical dimensions and applied voltages, while still maintaining constant the shape of the electric-field pattern. Thus two-dimensional effects are kept under control even though the intensity of the field is allowed to increase. The resulting design flexibility allows the design of FET's with quarter-micrometer channel length to be made, for either room temperature or liquid-nitrogen temperature. The physical limitations of the scaling theory are then investigated in detail, leading to the conclusion that the limiting FET performances are not reached at the 0.25-µm channel length. Further improvements are possible in the future, provided certain technology breakthroughs are achieved.

Current transport in Schottky diodes is examined by solving the Boltzmann equation within the semiconductor. The effect of the metal is taken into account by imposing suitable boundary conditions to the above solution, and it is shown that the usual drift‐diffusion equation is not correct in the vicinity of the junction. Generalized transport equations are derived, and the current‐voltage characteristic is determined. A thermionic‐emission velocity twice as large as that derived in the T‐D theory is obtained, in accordance with a previous Monte Carlo simulation of the M‐S device.

Electron dynamics in silicon is investigated by means of improved momentum- and energy-balance equations including particle diffusion and heat flux. The resulting system of partial differential equations is numerically solved in a variety of field configurations including strong discontinuities, in order to enhance velocity overshoot effects. It is found that diffusion, usually neglected in previous studies, plays a major role, and considerably modifies the features of the velocity vs distance curve, leading to an increase of the carrier drift velocity in the low-field region, i.e. before experiencing the effect of the strong field. In addition, it is found that, in order to take full advantage of velocity overshoot effects in MOSFET's, a structure must be designed having the strongest possible field at the source-end of the channel, where carrier density is controlled by the gate.

The range and straggle of boron in KTFR are measured and found to be in excellent agreement with the theoretical values of Lindhard, Scharff and Schiott for projected range, R p , and straggling, Δ R p , (depth of peak and width of assumed Gaussian distribution) if an empirical correction is made to the electronic stopping power using the results of Ormrod and Duckworth. A probable deviation from a Gaussian distribution approximation is found in this experiment for the fraction of ions which penetrate shallowly ( x 《 R p , where x is the depth) but a good fit to the Gaussian assumption is seen for the deeply penetrating ions ( x 《 R p ). From the results, curves are constructed which give the thickness of KTFR necessary to stop a given fraction of incident ions so a photoresist thickness sufficient for masking purposes may be chosen.

The spreading resistance due to current crowding at the end-points of an FET channel is investigated. An analytic expression is derived giving this resistance as function of a few parameters. Two-dimensional numerical simulations, using finite-element techniques, confirm the accuracy of the simple analytical approach. For short channel devices the current crowding effect is found to give a nonnegligible contribution to the total source resistance. In order to optimize the FET performance, the geometry and conductivity of the source/drain regions must be carefully designed, trading off short channel effect and transconductance degradation.

A new method for the analysis of the high‐ and low‐frequency MOS C–V characteristics is proposed to determine the interface‐state density in the presence of statistical fluctuations of the surface potential. This method provides simultaneously the mean oxide charge and the variance of its Gaussian distribution.

Infectious endocarditis involving intracardiac prosthesis continues to be a common and dangerous complication of cardiac surgery, but the incidence and severity of the disease are probably underestimated in this country. 14 cases of prosthetic endocarditis observed in a 5 years period are reported in this paper. 7 were early cases and 7 late cases. Etiology, clinical course, and pathological lesions, were similar in the two groups, while mortality was higher in the first group (100%) than in the second (85%). Delay in the diagnosis, negative blood cultures in a third of all cases, and doubts in indications for surgery, were the main reasons for a higher mortality than usual. Only one patient was reoperated on, unsuccessfully, but no patient who had not improved after two weeks of correct antibiotic treatment, survived; such cases, especially if diagnosis has been late and/or blood cultures negative, should be reoperated on without delay. The greatest efforts should be made in the prevention of the disease and, in particular, pre-operative and late post-operative prophylaxis should become more rigid and systematic.

The electrical properties of polycrystalline silicon resistors in the non-linear regime are examined. The adopted model is based on a thermionic-diffusion transport mechanism, and assumes charge trapping at the grain boundary. The theory closely fits experimental I-V characteristics, and allows for the determination of the average grain size.

A case of the Holt-Oram syndrome is described, that is, the association of a skeletal anomaly of the superior limbs, with a constant involvment of the thumbs, and an ostium secundum defect with a prolonged P-R interval. The case presented has the peculiarity of being sporadic, while from the literature a familiarity is almost constant via an autosomic dominant hereditary transmission. The more common syndromes were an alteration of the superior limbs is associated with a congenital cardiopathy are schematically underlined. Finally, the embriological reasons of this not rare association are briefly discussed.

A case of a 12 year old girl who presented tetralogy of Fallot (T. o F.) associated with Total Anomalous Pulmonary Venous Drainage (TAPVD) is described. This association is extremely rare, only four autopsied cases being reported in literature. TAPVD is associated with complex cyanotic congenital heart diseases in about one third of cases; asplenia or polysplenia are generally present. Double outlet right ventricle (DORV) is the commonest trunco-conal malformation associated with TAPVD; extremely rare is T. o F. Pulmonary outflow tract obstruction can avoid the haemodynamic consequences due to obstruction on the venous channel, or progression to pulmonary hypertension. The association of the two malformations can masquerade the clinical and Xray picture of TAPVD and make difficult the diagnosis during cardiac catheterization.

In a case of fetal heart failure caused by endocardial fibroelastosis, prenatal echocardiography clearly demonstrated a thickened endocardium. We therefore suggest that an abnormal endocardium may be detected in utero by ultrasound, thus representing an important clue in the differential diagnosis of fetal nonimmune hydrops and in the evaluation of pregnancies at risk for endocardial fibroelastosis.

The authors review eighty cases of pertrochanteric and subtrochanteric fractures of the femur treated surgically with the vitallium Jewett nail-plate between May 1975 and November 1977. They draw attention to the possibility offered by this fixation device of starting weight bearing at an early stage, a factor that is unanimously regarded as essential for rapid recovery in the elderly fracture patient. In fact, fifty-six patients commenced independent walking with some support within two weeks of operation. The fairly low incidence of complications (six in eighty operations), and the relative simplicity of the surgical technique, support the Jewett nail-plate as a suitable device for rigid fixation of trochanteric fractures of the neck of the femur.

The introduction of christiansen's prosthesis for the head of the femur has altered the balance between the two alternative surgical solutions for subcapital fractures of the neck of the femur (osteosynthesis and endoprosthesis). The writers report their results on the first forty-five patients treated surgically with the new prosthesis, illustrating its functional advantages as compared with the "single block" endoprosthesis. The risk of wear and tear of the acetabulum is practically abolished.

The frequency of operations on the knee, particularly on the antero-medial aspect, involving damage or section of sensory nerves, prompted the authors to carry out an accurate research on the innervation of this region. This revealed many variations of the normal anatomy which have been classified and illustrated in detail.