{"title":"具有空腔传感的大容量低温存储器","authors":"L. Burns, D. A. Christiansen, R. Gange","doi":"10.1145/1463822.1463833","DOIUrl":null,"url":null,"abstract":"This paper describes a cryoelectric high speed Continuous Sheet Memory (CSM) with 16,384 bit locations and complete cryotron addressing matrices. This memory is of the coincident current, random access type. The entire memory plane is fabricated via vacuum deposition techniques and is contained on a 2 inch by 2 inch substrate. In addition to a planar density of 10,000 bits per square inch, and complete cryotron decoding trees, the memory contains a novel sense structure in the form of a geometrical cavity. In addition to other advantages, the new cavity sense eliminates the major alignment problems. In one type of organization, each substrate or memory plane may be regarded as comprising the nth bit of 16,384 words; all n bits of a word may be read out in a parallel manner. Thus a word length is equal to the number of memory planes. This number of planes is at the discretion of the designer. The continuous sheet memory structure with decoding matrices and cavity sense offers a potential means toward the physical realization of a high speed random access memory of a capacity beyond a billion bits with present day technology. Since all present day computers of general utility are memory limited, the complexity of the mathematical operations performed by these computers is restricted by the size and/or access time of the memory. This restriction is indicative that large capacity high speed memories will play a major role in future computers. Many studies have been made of possible techniques that can be utilized for high speed, large capacity random access memories. These techniques have been in the areas of magnetics, semiconductors, and superconductivity. Microminiature magnetic cores, magnetic metal sheets, and deposited magnetic film show high promise toward the realization of a high speed random access medium capacity memory of about 107 bits, but do not appear to show promise toward large capacity memories of the order of 109 bits. Semiconductors, including transistors and tunnel diodes, also indicate a similar capability toward high speed memories of medium capacity. The present art of transistors is indicative that a very large array of the latter would be costly and cumbersome. Tunnel diodes are two terminal devices and have a relatively complicated memory cell structure. In addition such memories suffer an unavoidable delay due to the high amplification necessary because of signal attenuation; the latter is proportional to the number of words in the memory. A semiconductor device which has been proposed is the cryosar. This device is a two terminal element and awaits future development. In the area of superconductors, film cryotrons and persistors do not appear to offer the potential toward realizing a very large capacity memory of the order of 109 bits. The former has a relatively small bit density and registration difficulties; reasonably high uniformity is required and its relatively large power dissipation in large arrays places an upper bound on the number of cryotrons which may be utilized with present day technology. In addition, an inherent gain problem exists. A major disadvantage of the persister is the incomplete shielding which exists between the sense and drive lines. Thus arrays of only limited size can be made. A two-hole memory cell described by J. W. Crowe in 1957 was anticipated to hold high promise towards the realization of a high speed large capacity memory until it was found that the two-hole cell had an inherent uniformity problem due to the strong dependence of the critical field on the nature and edges of the holes. The continuous sheet memory eliminates the holes of the two-hole memory cell and with the cavity sense geometry represents perhaps the first realistic solution towards the physical realization of a practical high speed random access memory with a capacity in excess of a billion bits. This CSM with cavity sense eliminates the major registration problems encountered with present memory packing densities of 10,000 bits per square inch. In addition, the cryotron decoding matrices are an integral part of the CSM structure. Interrogation of the memory results in negligible power dissipation and no half-select noise has been observed. Present forms of the CSM have a signal-to-noise ratio of about 20:1 and low drive current requirements.","PeriodicalId":432708,"journal":{"name":"AFIPS '63 (Fall)","volume":"83 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1899-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"9","resultStr":"{\"title\":\"A large capacity cryoelectric memory with cavity sensing\",\"authors\":\"L. Burns, D. A. Christiansen, R. Gange\",\"doi\":\"10.1145/1463822.1463833\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper describes a cryoelectric high speed Continuous Sheet Memory (CSM) with 16,384 bit locations and complete cryotron addressing matrices. This memory is of the coincident current, random access type. The entire memory plane is fabricated via vacuum deposition techniques and is contained on a 2 inch by 2 inch substrate. In addition to a planar density of 10,000 bits per square inch, and complete cryotron decoding trees, the memory contains a novel sense structure in the form of a geometrical cavity. In addition to other advantages, the new cavity sense eliminates the major alignment problems. In one type of organization, each substrate or memory plane may be regarded as comprising the nth bit of 16,384 words; all n bits of a word may be read out in a parallel manner. Thus a word length is equal to the number of memory planes. This number of planes is at the discretion of the designer. The continuous sheet memory structure with decoding matrices and cavity sense offers a potential means toward the physical realization of a high speed random access memory of a capacity beyond a billion bits with present day technology. Since all present day computers of general utility are memory limited, the complexity of the mathematical operations performed by these computers is restricted by the size and/or access time of the memory. This restriction is indicative that large capacity high speed memories will play a major role in future computers. Many studies have been made of possible techniques that can be utilized for high speed, large capacity random access memories. These techniques have been in the areas of magnetics, semiconductors, and superconductivity. Microminiature magnetic cores, magnetic metal sheets, and deposited magnetic film show high promise toward the realization of a high speed random access medium capacity memory of about 107 bits, but do not appear to show promise toward large capacity memories of the order of 109 bits. Semiconductors, including transistors and tunnel diodes, also indicate a similar capability toward high speed memories of medium capacity. The present art of transistors is indicative that a very large array of the latter would be costly and cumbersome. Tunnel diodes are two terminal devices and have a relatively complicated memory cell structure. In addition such memories suffer an unavoidable delay due to the high amplification necessary because of signal attenuation; the latter is proportional to the number of words in the memory. A semiconductor device which has been proposed is the cryosar. This device is a two terminal element and awaits future development. In the area of superconductors, film cryotrons and persistors do not appear to offer the potential toward realizing a very large capacity memory of the order of 109 bits. The former has a relatively small bit density and registration difficulties; reasonably high uniformity is required and its relatively large power dissipation in large arrays places an upper bound on the number of cryotrons which may be utilized with present day technology. In addition, an inherent gain problem exists. A major disadvantage of the persister is the incomplete shielding which exists between the sense and drive lines. Thus arrays of only limited size can be made. A two-hole memory cell described by J. W. Crowe in 1957 was anticipated to hold high promise towards the realization of a high speed large capacity memory until it was found that the two-hole cell had an inherent uniformity problem due to the strong dependence of the critical field on the nature and edges of the holes. The continuous sheet memory eliminates the holes of the two-hole memory cell and with the cavity sense geometry represents perhaps the first realistic solution towards the physical realization of a practical high speed random access memory with a capacity in excess of a billion bits. This CSM with cavity sense eliminates the major registration problems encountered with present memory packing densities of 10,000 bits per square inch. In addition, the cryotron decoding matrices are an integral part of the CSM structure. Interrogation of the memory results in negligible power dissipation and no half-select noise has been observed. Present forms of the CSM have a signal-to-noise ratio of about 20:1 and low drive current requirements.\",\"PeriodicalId\":432708,\"journal\":{\"name\":\"AFIPS '63 (Fall)\",\"volume\":\"83 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"1899-12-30\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"9\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"AFIPS '63 (Fall)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/1463822.1463833\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"AFIPS '63 (Fall)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/1463822.1463833","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 9
摘要
本文介绍了一种具有16,384位的超低温高速连续片存储器(CSM)和完整的超低温寻址矩阵。该存储器为同步电流、随机存取类型。整个存储平面是通过真空沉积技术制造的,并包含在2英寸× 2英寸的基板上。除了每平方英寸10000比特的平面密度和完整的低温加速器解码树外,该存储器还包含一种几何空腔形式的新颖感知结构。除了其他优点,新的腔感消除了主要的对准问题。在一种类型的组织中,每个基板或存储平面可被视为包含16,384个字的第n位;一个字的所有n位可以以并行的方式读出。因此,一个字的长度等于存储平面的数量。飞机的数量由设计师自行决定。具有解码矩阵和空腔感测的连续片存储器结构为在现有技术条件下实现容量超过10亿比特的高速随机存取存储器提供了一种潜在的手段。由于目前所有的通用计算机都是内存有限的,这些计算机所执行的数学运算的复杂性受到内存大小和/或访问时间的限制。这一限制表明大容量高速存储器将在未来的计算机中发挥重要作用。人们对高速、大容量随机存取存储器的可能技术进行了许多研究。这些技术在磁学、半导体和超导等领域都有应用。微微型磁芯、磁性金属片和沉积磁膜对实现约107位的高速随机存取中等容量存储器大有希望,但对109位的大容量存储器似乎没有希望。包括晶体管和隧道二极管在内的半导体也显示出类似的中等容量高速存储器的能力。目前的晶体管技术表明,后者的一个非常大的阵列将是昂贵和笨重的。隧道二极管是两个终端器件,具有相对复杂的存储单元结构。此外,由于信号衰减所必需的高放大,这种存储器遭受不可避免的延迟;后者与记忆中的单词数量成正比。提出的一种半导体器件是低温星。该装置为双端元件,有待进一步开发。在超导体领域,薄膜低温管和持久管似乎不具备实现109位数量级的超大容量存储器的潜力。前者位密度相对较小,配准困难;需要相当高的均匀性,并且在大型阵列中其相对较大的功率耗散使现有技术可以使用的冷冻管数量的上限。此外,还存在固有增益问题。持续电阻的一个主要缺点是存在于感应线和驱动线之间的不完全屏蔽。因此,只能制作有限大小的数组。J. W. Crowe在1957年描述的双孔存储单元被认为有望实现高速大容量存储,直到发现双孔存储单元存在固有的均匀性问题,这是由于临界场对孔的性质和边缘的强烈依赖。连续片存储器消除了双孔存储器单元的孔,并且具有腔感几何结构,可能是实现容量超过10亿比特的实用高速随机存取存储器的第一个现实解决方案。这种具有腔感的CSM消除了目前存储器封装密度为每平方英寸10,000位时遇到的主要配准问题。此外,低温加速器解码矩阵是CSM结构的重要组成部分。对存储器的询问可以忽略功耗,并且没有观察到半选择噪声。目前形式的CSM具有约20:1的信噪比和低驱动电流要求。
A large capacity cryoelectric memory with cavity sensing
This paper describes a cryoelectric high speed Continuous Sheet Memory (CSM) with 16,384 bit locations and complete cryotron addressing matrices. This memory is of the coincident current, random access type. The entire memory plane is fabricated via vacuum deposition techniques and is contained on a 2 inch by 2 inch substrate. In addition to a planar density of 10,000 bits per square inch, and complete cryotron decoding trees, the memory contains a novel sense structure in the form of a geometrical cavity. In addition to other advantages, the new cavity sense eliminates the major alignment problems. In one type of organization, each substrate or memory plane may be regarded as comprising the nth bit of 16,384 words; all n bits of a word may be read out in a parallel manner. Thus a word length is equal to the number of memory planes. This number of planes is at the discretion of the designer. The continuous sheet memory structure with decoding matrices and cavity sense offers a potential means toward the physical realization of a high speed random access memory of a capacity beyond a billion bits with present day technology. Since all present day computers of general utility are memory limited, the complexity of the mathematical operations performed by these computers is restricted by the size and/or access time of the memory. This restriction is indicative that large capacity high speed memories will play a major role in future computers. Many studies have been made of possible techniques that can be utilized for high speed, large capacity random access memories. These techniques have been in the areas of magnetics, semiconductors, and superconductivity. Microminiature magnetic cores, magnetic metal sheets, and deposited magnetic film show high promise toward the realization of a high speed random access medium capacity memory of about 107 bits, but do not appear to show promise toward large capacity memories of the order of 109 bits. Semiconductors, including transistors and tunnel diodes, also indicate a similar capability toward high speed memories of medium capacity. The present art of transistors is indicative that a very large array of the latter would be costly and cumbersome. Tunnel diodes are two terminal devices and have a relatively complicated memory cell structure. In addition such memories suffer an unavoidable delay due to the high amplification necessary because of signal attenuation; the latter is proportional to the number of words in the memory. A semiconductor device which has been proposed is the cryosar. This device is a two terminal element and awaits future development. In the area of superconductors, film cryotrons and persistors do not appear to offer the potential toward realizing a very large capacity memory of the order of 109 bits. The former has a relatively small bit density and registration difficulties; reasonably high uniformity is required and its relatively large power dissipation in large arrays places an upper bound on the number of cryotrons which may be utilized with present day technology. In addition, an inherent gain problem exists. A major disadvantage of the persister is the incomplete shielding which exists between the sense and drive lines. Thus arrays of only limited size can be made. A two-hole memory cell described by J. W. Crowe in 1957 was anticipated to hold high promise towards the realization of a high speed large capacity memory until it was found that the two-hole cell had an inherent uniformity problem due to the strong dependence of the critical field on the nature and edges of the holes. The continuous sheet memory eliminates the holes of the two-hole memory cell and with the cavity sense geometry represents perhaps the first realistic solution towards the physical realization of a practical high speed random access memory with a capacity in excess of a billion bits. This CSM with cavity sense eliminates the major registration problems encountered with present memory packing densities of 10,000 bits per square inch. In addition, the cryotron decoding matrices are an integral part of the CSM structure. Interrogation of the memory results in negligible power dissipation and no half-select noise has been observed. Present forms of the CSM have a signal-to-noise ratio of about 20:1 and low drive current requirements.