Misplaced Pages

#657342

73-564: On September 5, 1990, IBM published a group of hardware and software announcements, two of which included overviews of three announcements: Despite the fact that IBM mentioned the 9000 family first in some of the day's announcements, it was clear "by the end of the day" that it was "for System/390," although it was a shortened name, S/390 , that was placed on some of the actual "boxes" later shipped. The ES/9000 include rack-mounted models, free standing air cooled models and water cooled models. The low end models were substantially less expensive than

146-462: A common region that minority carriers can move through. A PNP BJT will function like two diodes that share an N-type cathode region, and the NPN like two diodes sharing a P-type anode region. Connecting two diodes with wires will not make a BJT, since minority carriers will not be able to get from one p–n junction to the other through the wire. Both types of BJT function by letting a small current input to

219-527: A data center environment. They had a disruptive effect on the market, allowing customers to provide internal IBM computing services at a cost point lower than commercial time-sharing services. All 4300 processors used a 3278-2A, 3279-C or 3205 display console rather than a 3210 or 3215 keyboard/printer console. Each model - 4331, 4341, 4361, and 4381 - had various sub-models, such as the 4341 model 1 (or 4341-1) and 4341 model 2 (4341-2). The 4381-13 through 4381-24 (announced in 1987) were entry-level machines for

292-406: A few hundred millivolts) biases. For example, in the typical grounded-emitter configuration of an NPN BJT used as a pulldown switch in digital logic, the "off" state never involves a reverse-biased junction because the base voltage never goes below ground; nevertheless the forward bias is close enough to zero that essentially no current flows, so this end of the forward active region can be regarded as

365-401: A junction between two regions of different charge carrier concentration. The regions of a BJT are called emitter , base , and collector . A discrete transistor has three leads for connection to these regions. Typically, the emitter region is heavily doped compared to the other two layers, and the collector is doped more lightly (typically ten times lighter ) than the base. By design, most of

438-506: A more positive potential than the n-doped side, and the base–collector junction is reverse biased . When forward bias is applied to the base–emitter junction, the equilibrium between the thermally generated carriers and the repelling electric field of the emitter depletion region is disturbed. This allows thermally excited carriers (electrons in NPNs, holes in PNPs) to inject from the emitter into

511-407: A small current injected at one of its terminals to control a much larger current between the remaining two terminals, making the device capable of amplification or switching . BJTs use two p–n junctions between two semiconductor types, n-type and p-type, which are regions in a single crystal of material. The junctions can be made in several different ways, such as changing the doping of

584-583: A standard feature of the ES/9000 processors whereby IBM's Processor Resource/Systems Manager (PR/SM) hypervisor allows different operating systems to run concurrently in separate logical partitions (LPARs), with a high degree of isolation. Initially 7 partitions per a disconnected side were supported. In December 1992 the LPAR capacity of the H2 (520-based) models was increased to 10 per a disconnected side. For example,

657-423: A thin p-doped region, and a PNP transistor comprises two semiconductor junctions that share a thin n-doped region. N-type means doped with impurities (such as phosphorus or arsenic ) that provide mobile electrons, while p-type means doped with impurities (such as boron ) that provide holes that readily accept electrons. Charge flow in a BJT is due to diffusion of charge carriers (electrons and holes) across

730-402: A two-processor model 660 could now support up to 20 partitions instead of 14, if the two sides (each with one processor) are electrically isolated. This was introduced as part of IBM's moving towards "lights-out" operation and increased control of multiple system configurations. Launched in 1994 first as the "Parallel Transaction Server" (alongside the 9673 "Parallel Query Server"), subsumed by

803-563: A very low cost. Bipolar transistor integrated circuits were the main active devices of a generation of mainframe and minicomputers , but most computer systems now use Complementary metal–oxide–semiconductor ( CMOS ) integrated circuits relying on the field-effect transistor (FET). Bipolar transistors are still used for amplification of signals, switching, and in mixed-signal integrated circuits using BiCMOS . Specialized types are used for high voltage switches, for radio-frequency (RF) amplifiers, or for switching high currents. By convention,

SECTION 10

#1732779605658

876-551: A year later than the models with older technology. Later these new technologies were used in models 520, 640, 660, 740 and 860. All three lines got additions and upgrades until 1993–1994. In February 1993 an 8-processor 141 MHz (7.1 ns ) model 982 became available, with models 972, 962, 952, 942, 941, 831, 822, 821 and 711 following in March. These models, codenamed H5, had double the L2 cache and 30% higher per-processor performance than

949-724: Is an improvement of the BJT that can handle signals of very high frequencies up to several hundred GHz . It is common in modern ultrafast circuits, mostly RF systems. Two commonly used HBTs are silicon–germanium and aluminum gallium arsenide, though a wide variety of semiconductors may be used for the HBT structure. HBT structures are usually grown by epitaxy techniques like MOCVD and MBE . Bipolar transistors have four distinct regions of operation, defined by BJT junction biases: Although these regions are well defined for sufficiently large applied voltage, they overlap somewhat for small (less than

1022-411: Is called active mode, the base–emitter voltage V BE {\displaystyle V_{\text{BE}}} and collector–base voltage V CB {\displaystyle V_{\text{CB}}} are positive, forward biasing the emitter–base junction and reverse-biasing the collector–base junction. In this mode, electrons are injected from the forward biased n-type emitter region into

1095-409: Is made from lightly doped, high-resistivity material. The collector surrounds the emitter region, making it almost impossible for the electrons injected into the base region to escape without being collected, thus making the resulting value of α very close to unity, and so, giving the transistor a large β. A cross-section view of a BJT indicates that the collector–base junction has a much larger area than

1168-447: Is related to V BE {\displaystyle V_{\text{BE}}} exponentially. At room temperature , an increase in V BE {\displaystyle V_{\text{BE}}} by approximately 60 mV increases the emitter current by a factor of 10. Because the base current is approximately proportional to the collector and emitter currents, they vary in the same way. The bipolar point-contact transistor

1241-577: Is reverse biased in normal operation. The reason the emitter is heavily doped is to increase the emitter injection efficiency: the ratio of carriers injected by the emitter to those injected by the base. For high current gain, most of the carriers injected into the emitter–base junction must come from the emitter. The low-performance "lateral" bipolar transistors sometimes used in CMOS processes are sometimes designed symmetrically, that is, with no difference between forward and backward operation. Small changes in

1314-408: Is usually 100 or more, but robust circuit designs do not depend on the exact value (for example see op-amp ). The value of this gain for DC signals is referred to as h FE {\displaystyle h_{\text{FE}}} , and the value of this gain for small signals is referred to as h fe {\displaystyle h_{\text{fe}}} . That is, when a small change in

1387-571: The 370-XA architecture. They were positioned between the IBM 9370 and IBM 3090 in performance at the time of announcement. The 4381-3, 4381-14, 4381-24 and 4381-92 are dual-CPU models. Other models included 1, 2, 11, 12, 13, 21, 22, 23, 90 and 91. The IBM 4321 was announced on 18 November 1981. The IBM 4331 (and the 4341) were announced on 30 January 1979. It came with an integrated adapter that permitted attaching up to 16 of two newly introduced direct-access storage devices (DASD): The 4331

1460-520: The Ebers–Moll model ) is required. The voltage-control model requires an exponential function to be taken into account, but when it is linearized such that the transistor can be modeled as a transconductance, as in the Ebers–Moll model, design for circuits such as differential amplifiers again becomes a mostly linear problem, so the voltage-control view is often preferred. For translinear circuits , in which

1533-416: The emitter region, the base region and the collector region. These regions are, respectively, p type, n type and p type in a PNP transistor, and n type, p type and n type in an NPN transistor. Each semiconductor region is connected to a terminal, appropriately labeled: emitter (E), base (B) and collector (C). The base is physically located between the emitter and the collector and

SECTION 20

#1732779605658

1606-459: The "Parallel Enterprise Server" launched later in the year, the six generations of the IBM 9672 machines transitioned IBM's mainframes fully to CMOS microprocessors, as by a strategic decision no more ES/9000 ( bipolar -based except the 9221) models would be released after 1994. The initial generations of 9672 were slower than the largest ES/9000 sold in parallel, but the fifth and sixth generations were

1679-552: The 3090 or 4381 previously needed to run MVS/ESA , and could also run VM/ESA and VSE/ESA , which IBM announced at the same time. IBM periodically added named features to ESA/390 in conjunction with new processors; the ESA/390 Principles of Operation manual identifies them only by name, not by the processors supporting them. Machines supporting the architecture were sold under the brand System/390 (S/390) from September 1990. The 9672 implementations of System/390 were

1752-411: The 4300 series as well as other System/370-compatible processors. For the 4321 and 4331: Bipolar junction transistor A bipolar junction transistor ( BJT ) is a type of transistor that uses both electrons and electron holes as charge carriers . In contrast, a unipolar transistor, such as a field-effect transistor (FET), uses only one kind of charge carrier. A bipolar transistor allows

1825-442: The 69 MHz (14.5 ns) maximum frequency and thus with unchanged performance. Those models' main difference from the 3090-J was the optional addition of ESCON , Sysplex and Integrated Cryptographic Feature. Only the models 900 and 820 had an all-new design (codenamed H2), featuring private split I+D 128+128 KB L1 caches and a shared 4 MB L2 cache (2 MB per side) with 11-cycle latency, more direct interconnects between

1898-519: The 9221. The 9221 models 120, 130 and 150 were initially available only with the "System/370 Base Option"; the "ESA Option" shipped in July 1991. The 9221 processors were made of VLSI CMOS chips designed in Böblingen , Germany, whence the 9672 line later originated. The lower 6 of the 8 water-cooled models (codenamed H0) were immediately available, but used the same processor as the 3090-J, still at

1971-542: The 9672 G5, became available in September 1999, featuring PCI buses. The S/390 Integrated Server , an even lower-end S/390 system than Multiprise, shipped by the end of 1998. It emerged from a line of S/390-compatibility/coprocessor cards for PCs, but is a true S/390 system capable of server duties, having relegated the Pentium II to the role of an I/O coprocessor. It was the first S/390 server to support PCI. It had

2044-401: The BJT collector current is due to the flow of charge carriers injected from a heavily doped emitter into the base where they are minority carriers (electrons in NPNs, holes in PNPs) that diffuse toward the collector, so BJTs are classified as minority-carrier devices . In typical operation, the base–emitter junction is forward biased , which means that the p-doped side of the junction is at

2117-635: The DEC Alphas into early 1999. The G5 also added support for the IEEE 754 floating-point formats. The thousandth G5 system shipped less than 100 days after the manufacturing began; the greatest ramping of production in S/390's history. In late May 1999 the G6 arrived featuring copper interconnects , raising the frequency to 637 MHz, higher than the fastest DEC machines at the time. In September 1996 IBM launched

2190-417: The ES/9000 in three form factors; the water-cooled 9021 to succeed the IBM 3090 , and the air-cooled standalone 9121 and rack-mounted 9221 to succeed the IBM 4381 and 9370 respectively. The largest announced model had a 100-fold performance over the smallest model, and the clock frequency ranged from 67-111 MHz (15-9  ns ) in the 9021 and 67 MHz in the 9121 to 26-33 MHz (38-30 ns) in

2263-450: The Ebers–Moll model: The base internal current is mainly by diffusion (see Fick's law ) and where The α {\displaystyle \alpha } and forward β {\displaystyle \beta } parameters are as described previously. A reverse β {\displaystyle \beta } is sometimes included in the model. The unapproximated Ebers–Moll equations used to describe

IBM System/390 - Misplaced Pages Continue

2336-593: The Extended Control Program Support:MVS (ECPS:MVS) option, a subset of System/370 extended facility. On 20 October 1982, IBM announced a new entry-level 4341 model, Model Group 9 and a new top-of-the-line 4341, Model Group 12. Model Group 12 included the Dual Address Space (DAS) facility. The 4341 was withdrawn on 11 February 1986. The IBM 4361 Model Groups 4 and 5 were announced on 15 September 1983. Model Group 3

2409-502: The G3 at silicon process parity, but it suffered a 23% IPC reduction from the G3. The initial G4-based models became available in June 1997, but it wasn't until the 370 MHz model RY5 (with a "Modular Cooling Unit") became available at the end of the year that a 9672 would almost match the 141 MHz model 9X2's performance. At 370 MHz it was the second-highest clocked microprocessor at

2482-572: The H2 line, and added a hardware data compression . The compression was also included in the new, 50% faster models of the 9121. In April 1994, alongside the CMOS -based new 9672 series and improved 9221 models (with 40% faster cycle time and data compression), IBM announced also their ultimate bipolar model, the 10-processor model 9X2 rated at 468 MIPS, to become available in October. Previously available only on IBM 3090 , Logical Partitions (LPARs) are

2555-527: The S/390 Multiprise 2000, positioned below the 9672. It used the same technology as the 9672 G3, but it fit half as many processors (up to five) and its off-chip caches were smaller. The 9672 G3 and the Multiprise 2000 were the last versions to support pre-XA System/370 mode. In October 1997 models of Multiprise 2000 with an 11% higher performance were launched. The Multiprise 3000 , based on

2628-427: The absorption of photons , and handles the dynamics of turn-off, or recovery time, which depends on charge in the base region recombining. However, because base charge is not a signal that is visible at the terminals, the current- and voltage-control views are generally used in circuit design and analysis. In analog circuit design, the current-control view is sometimes used because it is approximately linear. That is,

2701-400: The base and reach the collector is a measure of the BJT efficiency. The heavy doping of the emitter region and light doping of the base region causes many more electrons to be injected from the emitter into the base than holes to be injected from the base into the emitter. A thin and lightly doped base region means that most of the minority carriers that are injected into the base will diffuse to

2774-465: The base control an amplified output from the collector. The result is that the BJT makes a good switch that is controlled by its base input. The BJT also makes a good amplifier, since it can multiply a weak input signal to about 100 times its original strength. Networks of BJTs are used to make powerful amplifiers with many different applications. In the discussion below, focus is on the NPN BJT. In what

2847-427: The base reduce the BJT gain. Another useful characteristic is the common-base current gain , α F . The common-base current gain is approximately the gain of current from emitter to collector in the forward-active region. This ratio usually has a value close to unity; between 0.980 and 0.998. It is less than unity due to recombination of charge carriers as they cross the base region. Alpha and beta are related by

2920-400: The base region. These carriers create a diffusion current through the base from the region of high concentration near the emitter toward the region of low concentration near the collector. To minimize the fraction of carriers that recombine before reaching the collector–base junction, the transistor's base region must be thin enough that carriers can diffuse across it in much less time than

2993-434: The base–emitter junction and recombination in the base). In many designs beta is assumed high enough so that base current has a negligible effect on the circuit. In some circuits (generally switching circuits), sufficient base current is supplied so that even the lowest beta value a particular device may have will still allow the required collector current to flow. BJTs consists of three differently doped semiconductor regions:

IBM System/390 - Misplaced Pages Continue

3066-467: The characteristics allows designs to be created following a logical process. Bipolar transistors, and particularly power transistors, have long base-storage times when they are driven into saturation; the base storage limits turn-off time in switching applications. A Baker clamp can prevent the transistor from heavily saturating, which reduces the amount of charge stored in the base and thus improves switching time. The proportion of carriers able to cross

3139-507: The collector and not recombine. The common-emitter current gain is represented by β F or the h -parameter h FE ; it is approximately the ratio of the collector's direct current to the base's direct current in forward-active region. (The F subscript is used to indicate the forward-active mode of operation.) It is typically greater than 50 for small-signal transistors, but can be smaller in transistors designed for high-power applications. Both injection efficiency and recombination in

3212-405: The collector current is approximately β F {\displaystyle \beta _{\text{F}}} times the base current. Some basic circuits can be designed by assuming that the base–emitter voltage is approximately constant and that collector current is β times the base current. However, to accurately and reliably design production BJT circuits, the voltage-control model (e.g.

3285-458: The collector–base depletion region, are swept into the collector by the electric field in the depletion region. The thin shared base and asymmetric collector–emitter doping are what differentiates a bipolar transistor from two separate diodes connected in series. The collector–emitter current can be viewed as being controlled by the base–emitter current (current control), or by the base–emitter voltage (voltage control). These views are related by

3358-466: The collector–base voltage, for example, causes a greater reverse bias across the collector–base junction, increasing the collector–base depletion region width, and decreasing the width of the base. This variation in base width often is called the Early effect after its discoverer James M. Early . Narrowing of the base width has two consequences: Both factors increase the collector or "output" current of

3431-465: The conventional direction, but labels for the movement of holes and electrons show their actual direction inside the transistor. The arrow on the symbol for bipolar transistors indicates the p–n junction between base and emitter and points in the direction in which conventional current travels. BJTs exist as PNP and NPN types, based on the doping types of the three main terminal regions. An NPN transistor comprises two semiconductor junctions that share

3504-508: The currents occurs, and sufficient time has passed for the new condition to reach a steady state h fe {\displaystyle h_{\text{fe}}} is the ratio of the change in collector current to the change in base current. The symbol β {\displaystyle \beta } is used for both h FE {\displaystyle h_{\text{FE}}} and h fe {\displaystyle h_{\text{fe}}} . The emitter current

3577-402: The current–voltage relation of the base–emitter junction, which is the usual exponential current–voltage curve of a p–n junction (diode). The explanation for collector current is the concentration gradient of minority carriers in the base region. Due to low-level injection (in which there are much fewer excess carriers than normal majority carriers) the ambipolar transport rates (in which

3650-476: The cutoff region. The diagram shows a schematic representation of an NPN transistor connected to two voltage sources. (The same description applies to a PNP transistor with reversed directions of current flow and applied voltage.) This applied voltage causes the lower p–n junction to become forward biased, allowing a flow of electrons from the emitter into the base. In active mode, the electric field existing between base and collector (caused by V CE ) will cause

3723-672: The design of digital integrated circuits. The incidental low performance BJTs inherent in CMOS ICs, however, are often utilized as bandgap voltage reference , silicon bandgap temperature sensor and to handle electrostatic discharge . The germanium transistor was more common in the 1950s and 1960s but has a greater tendency to exhibit thermal runaway . Since germanium p-n junctions have a lower forward bias than silicon, germanium transistors turn on at lower voltage. Various methods of manufacturing bipolar transistors were developed. BJTs can be thought of as two diodes (p–n junctions) sharing

SECTION 50

#1732779605658

3796-511: The direction of current on diagrams is shown as the direction that a positive charge would move. This is called conventional current . However, current in metal conductors is generally due to the flow of electrons. Because electrons carry a negative charge, they move in the direction opposite to conventional current. On the other hand, inside a bipolar transistor, currents can be composed of both positively charged holes and negatively charged electrons. In this article, current arrows are shown in

3869-416: The emitter–base junction. The bipolar junction transistor, unlike other transistors, is usually not a symmetrical device. This means that interchanging the collector and the emitter makes the transistor leave the forward active mode and start to operate in reverse mode. Because the transistor's internal structure is usually optimized for forward-mode operation, interchanging the collector and the emitter makes

3942-479: The excess majority and minority carriers flow at the same rate) is in effect determined by the excess minority carriers. Detailed transistor models of transistor action, such as the Gummel–Poon model , account for the distribution of this charge explicitly to explain transistor behaviour more exactly. The charge-control view easily handles phototransistors , where minority carriers in the base region are created by

4015-414: The exponential I–V curve is key to the operation, the transistors are usually modeled as voltage-controlled current sources whose transconductance is proportional to their collector current. In general, transistor-level circuit analysis is performed using SPICE or a comparable analog-circuit simulator, so mathematical model complexity is usually not of much concern to the designer, but a simplified view of

4088-518: The first high-end IBM mainframe architecture implemented first with CMOS CPU electronics rather than the traditional bipolar logic. The IBM z13 was the last z Systems server to support running an operating system in ESA/390 architecture mode. However, all 24-bit and 31-bit problem-state application programs originally written to run on the ESA/390 architecture readily run unaffected by this change. Eighteen models were announced September 5, 1990 for

4161-410: The following identities: Beta is a convenient figure of merit to describe the performance of a bipolar transistor, but is not a fundamental physical property of the device. Bipolar transistors can be considered voltage-controlled devices (fundamentally the collector current is controlled by the base–emitter voltage; the base current could be considered a defect and is controlled by the characteristics of

4234-400: The majority of these electrons to cross the upper p–n junction into the collector to form the collector current I C . The remainder of the electrons recombine with holes, the majority carriers in the base, making a current through the base connection to form the base current, I B . As shown in the diagram, the emitter current, I E , is the total transistor current, which is the sum of

4307-427: The most powerful and capable ESA/390 machines built by IBM. In the course of the generations, CPUs added more instructions and increased performance. The first three generations (G1 to G3) focused on low cost. The 4th generation was aimed at matching the performance of the last bipolar model, the 9021-9X2. It was decided to be accomplished by pursuing high clock frequencies. The G4 could reach 70% higher frequency than

4380-469: The other terminal currents, (i.e. I E  =  I B  +  I C ). In the diagram, the arrows representing current point in the direction of conventional current – the flow of electrons is in the opposite direction of the arrows because electrons carry negative electric charge . In active mode, the ratio of the collector current to the base current is called the DC current gain . This gain

4453-470: The p-type base where they diffuse as minority carriers to the reverse-biased n-type collector and are swept away by the electric field in the reverse-biased collector–base junction. For an illustration of forward and reverse bias, see semiconductor diodes . In 1954, Jewell James Ebers and John L. Moll introduced their mathematical model of transistor currents: The DC emitter and collector currents in active mode are well modeled by an approximation to

SECTION 60

#1732779605658

4526-542: The processors, multi-level TLBs , branch target buffer and 111 MHz (9  ns ) clock frequency. These were the first models with out-of-order execution since the System/370-195 of 1973. However unlike the old S/360-91 -derived systems, the models 900 and 820 had full out-of-order execution for both integer and floating-point units, with precise exception handling , and a fully superscalar pipeline. Models 820 and 900 shipped to customers in September 1991,

4599-490: The same performance and 256 MB maximum memory capacity as the 7 years older low-end 9221 model 170. From 1997 IBM also offered a "S/390 Application StarterPak", intended as a software development kit for developing and testing mainframe software. IBM 4300#IBM 4381 The IBM 4300 series are mid-range systems compatible with System/370 that were sold from 1979 through 1992. They featured modest electrical and cooling requirements, and thus did not require

4672-506: The semiconductor material as it is grown, by depositing metal pellets to form alloy junctions, or by such methods as diffusion of n-type and p-type doping substances into the crystal. The superior predictability and performance of junction transistors quickly displaced the original point-contact transistor . Diffused transistors, along with other components, are elements of integrated circuits for analog and digital functions. Hundreds of bipolar junction transistors can be made in one circuit at

4745-421: The semiconductor's minority-carrier lifetime. Having a lightly doped base ensures recombination rates are low. In particular, the thickness of the base must be much less than the diffusion length of the carriers. The collector–base junction is reverse-biased, and so negligible carrier injection occurs from the collector to the base, but carriers that are injected into the base from the emitter, and diffuse to reach

4818-417: The three currents in any operating region are given below. These equations are based on the transport model for a bipolar junction transistor. where As the collector–base voltage ( V CB = V CE − V BE {\displaystyle V_{\text{CB}}=V_{\text{CE}}-V_{\text{BE}}} ) varies, the collector–base depletion region varies in size. An increase in

4891-514: The time, after the Alpha 21164 of DEC . The execution units in each G4 processor are duplicated for the purpose of error detection and correction. Arriving in late September 1998, the G5 more than doubled the performance over any previous IBM mainframe, and restored IBM's performance lead that had been lost to Hitachi 's Skyline mainframes in 1995. The G5 operated at up to 500 MHz, again second only to

4964-428: The values of α and β in reverse operation much smaller than those in forward operation; often the α of the reverse mode is lower than 0.5. The lack of symmetry is primarily due to the doping ratios of the emitter and the collector. The emitter is heavily doped, while the collector is lightly doped, allowing a large reverse bias voltage to be applied before the collector–base junction breaks down. The collector–base junction

5037-670: The voltage applied across the base–emitter terminals cause the current between the emitter and the collector to change significantly. This effect can be used to amplify the input voltage or current. BJTs can be thought of as voltage-controlled current sources , but are more simply characterized as current-controlled current sources, or current amplifiers, due to the low impedance at the base. Early transistors were made from germanium but most modern BJTs are made from silicon . A significant minority are also now made from gallium arsenide , especially for very high speed applications (see HBT, below). The heterojunction bipolar transistor (HBT)

5110-475: Was announced the following year on 12 September 1984. Among the new/optional features for the 4361 were: While floating-point arithmetic has long been part of computing history, and was present in System/360 , this feature's advancement, conceptualization of which, as Karlsruhe Accurate Arithmetic , had been under development for decades, was implemented as an optional feature on the 4361. The 4361

5183-555: Was invented in December 1947 at the Bell Telephone Laboratories by John Bardeen and Walter Brattain under the direction of William Shockley . The junction version known as the bipolar junction transistor (BJT), invented by Shockley in 1948, was for three decades the device of choice in the design of discrete and integrated circuits . Nowadays, the use of the BJT has declined in favor of CMOS technology in

5256-470: Was withdrawn on 17 February 1987. The IBM 4381 had a greater longevity than any of the above systems. Model Groups 1 and 2 were announced Sep 15, 1983 and withdrawn on 11 February 1986. Model Group 3 was announced on 25 October 1984 and withdrawn on 11 February 1986. Model Groups 11, 12, 13 and 14 were announced on 11 February 1986. Model Groups 21, 22, 23 and 24 were announced on 19 May 1987 and withdrawn on 19 August 1992. New releases of: supported

5329-602: Was withdrawn on 18 November 1981. The IBM 4341 (and the 4331) were announced on 30 January 1979. Like the 4331, it came with an integrated adapter that permitted attaching up to 16 of the newly introduced IBM 3370 DASD. The 4341 did not support the much lower capacity IBM 3310. The 4341 Introduced the Extended Control Program Support:VM (ECPS:VM), Extended Control Program Support:VS1 (ECPS:VS1) and Extended Control Program Support:Virtual Storage Extended (ECPS:VSE) features. The 4341-2 introduced

#657342