Integrated Circuit Chip Size and Levels of Integration

Integrated Circuit Chip Size and Levels of Integration

In 1947, Prof. Bardeen and Prof. Brattain, the members of a research group at Bell Laboratories, invented first germanium transistor. Actually the team was attempting to make a device resembling a field effect transistor. Eventhough the basic principle behind the operation of MOSFET was proposed in 1935, it took long period to materialize this device. By the year of 1951, the Bipolar Junction Transistors (BJTs) and the Junction Field Effect Transistors (JFETs)s came in the market. Immediately in 1954, silicon took the place of germanium in BJTs. In late 1960, first time MOSFET was produced.

Meanwhile the interconnection of electric circuits with the help of the printed circuit assembly was under progress. In 1952, the scientist found the possibility of semiconductor integrated circuits. Finally in 1958 Jack Kilby of Texas Instruments

invented first integrated circuit. Shortly after this invention Prof. Robert Noyce and Gorden Moore of Fairchild semiconductor, USA, announced the planar process for silicon bipolar transistors which stepped ahead to the development of monolithic IC. After realizing the strength of the IC, it become the most demanded device. Then progressively inventions were made in increasing the number of active devices per chip by reducing device dimensions.

Depending upon the number of active devices per chip, there are different levels of integration as explained below.

When the active devices per chip are less than 100, then it is referred as Small Scale Integration (SSI). Most of the SSI chips use integrated resistors, diodes and bipolar transistors.

When the count of active devices per chip is between 100 to 1000, then it is referred as Medium Scale Integration (MSI). In most of the MSI chips, BJTs and enhancement mode MOSFETs are integrated.

In Large Scale Integration (LSI) ICs, the number of active devices per chip ranges between 1000 to 100,000. In general, LSI chips use MOS transistors; as it requires less number of steps for integration. Thus more number of components can be produced on the chip with MOS transistors than with the bipolar transistors.

When the active devices per chip are over hundreds of thousands, then it is referred as Very Large Scale Integration (VLSI). Almost all modem chips employ VLSI technique.

In general, from Small Scale Integration (SSI) level to the Very Large Scale Integration (VLSI) level, the device density has been increased tremendously. Also the chip area is increased from 1 mm (for SSI level) to 1 cm (for VLSI level). The variation in the chip area for SSI chip to VLSI chip is a shown in the Fig. 1.3.1.

Recently a new level of integration has been introduced which is known as Ultra Large Scale Integration (ULSI). In ULSI technique, more than one million active devices are integrated on a single chip. Pentium microprocessors use ULSI technology.

Table 1.3.1 gives the categories of ICs on the basis of number of active devices. Note that the number of active devices per chip is nothing but the number of transistors per chip. Sometimes an alternative parameter for transistor count per chip may be used. It is referred as gate count.

Review Question

1. Explain briefly about different levels of integration.