Online Postgraduate Courses for the Electronics Industry

Provided By The University of Bolton

• Home
• About Us
• Questions & Answers
• Courses
• Students
• Links

Useful Tools

Microelectronic Technologies and Applications

Unit 1: CMOS Digital Logic

This chapter reviews aspects of ASICs and the process involved in designing and fabricating them using the common CMOS process. This leads to the design and implementation of common logic cells found in most digital designs including combinational, sequential, and I/O cells.

Unit Contents

• 1.1 Introduction
  • 1.1.1 Types of ASICS
  • 1.1.2 Full Custom ASICS
  • 1.1.3 Standard Cell Based ASIC
  • 1.1.4 Gate Array ASIC
  • 1.1.5 Programmable Logic Devices
• 1.2 ASIC Design Flow
• 1.3 Case Study
• 1.4 ASIC Cost Comparisons 
• 1.5 Cell Libraries 
• 1.6 Websites of Interest 
• 1.7 CMOS Inverter 
• 1.8  CMOS Process 
• 1.9 CMOS Design Rules 
• 1.10  Logic Cells 
  • 1.10.1  Combinational 
  • 1.10.2  Sequential 
  • 1.10.3  Datapath Logic Cells 

  • ---

    1.10.4  Datapath Cells 
  • 1.10.5  I/O Cells 
  • 1.10.6  Compiled Cells 
• Self Assessment
• Interactive CMOS Inverter
• Excel Spreadsheet Exercise

1.1 Introduction

An Application Specific Integrated Circuit (ASIC) is one type of silicon integrated circuit (IC) or chip. Other types are memories (Flash, DRAM and SRAM), microprocessors and standard devices. In essence ASICs are devices made for a specific application such as a mobile phone.

Commercial ICs are typically enclosed or mounted or or in a package or substrate, which can be made from ceramic or plastic. The silicon chip itself is can range from 1mm square up to 100's of mm2 and contain billions of transistors.

Intel 30 Mar 2010 ... The die size is 684 mm2 with 2.3 billion transistors

The course text refers to the book

For more information on the development of ICs read the first section of chapter 1 of the text-book. (Application-specific Integrated Circuits by Michael John Smith, ISBN: 0201500221, Publisher: Addison Wesley available on line at the EDA Cafe web:http://www.edacafe.com/book/ASIC/CH01/CH01.php

The EDA Cafe's other offerings are also well worth perusing.

Alternatives would be:

• CMOS VLSI Design Weste/Harris

There are many types of ASICs so we will cover a limited set of them in bried.

1.1.1 Types of ASICs

All ASICs are fabricated on silicon wafers (typically 300mm-600mm in diameter) by building up layers with different semiconductor characteristics in order to produce transistors, interconnections, capacitors, inductors and resistors required. The process used to define certain areas is masking , which is a process akin to very high definition photographic printing. A large number of other stages are required to produce the ASIC. Typically the minimum feature size of the process is the length of the NMOS or PMOS gate. A 0.13um process will have a nominal minimum size for the FET transistor gate length of 0.13um, allowing for process offsets. Hence a 45nm process has a minimum gate length of 45nm. The wafers are then sawn into individual die and mounted in an IC package. In total the production of an ASIC takes typically 8-10 weeks.

A full custom chip is made specifically and uniquely for one application , is very expensive and only economic at very high volumes (millions/year).

To make ASICs economic at lower volumes the semi-custom concept was introduced where many applications share the same basic configuration of logic cells – it is only the final interconnect stage that is different to give the different chips. There are several types of semi- custom ASICs – in succeeding sections we will look briefly at each type.

1.1.2 Full custom ASICs

In full custom the designer essentially works at the transistor level designing sub- systems and systems to meet his specification precisely. The chip may have up to billions of transistors so very complex designs can be made. He may use predefined cells if they are available and suitable but, in general, all/ parts of the design have to be designed in this way. Therefore full custom is slow (even with the use of modern CAD tools) and typical design times for a large full custom chip can be several hundred man-years or more. It follows therefore that full custom designs are very expensive and the initial cost can only be recouped if the chip volumes are very high.

Whilst synthesis made the implementation of complex digital systems significantly easier as modern geometries have decreased the impact of the increasing track capacitance over the gate capacitance has meant that standard estimations methodologies were insufficient to ensure timing closure on modern systems with their high frequency performance. Hence physically knowledgable synthesis tools which also do the cell layout have been required.

However in analogue and mixed analogue/digital systems where the matching and electrical characteristics are critical and full custom is commonly used.

1.1.3 Standard cell based ASIC (or cell-based IC CBIC)

This approach uses pre-defined standard logic cells (gates, flip-flops etc) and larger blocks such as processors, which are pre-defined and characterised. The designer simply has to build up his design , which makes the whole process quicker and cheaper and makes designs implemented in this way economic at lower volumes (typically 1000k units/year) than full custom albeit at a loss of some flexibility for the designer.

Figure 1.1: Physical layout of a CBIC design.

(from Application-Specific Integrated Circuits (fig 1.2) by MJS Smith - permission requested)

1.1.4 Gate array ASIC (channelless array or sea of gates array)

In this case the transistors are arranged as a regular array on the ASIC and the designer has only to design the final interconnect pattern using transistors or logic cells in a cell library and powerful CAD tools. See Figure 1.2.

Figure 1.2: Gate array ASIC

(from Application-Specific Integrated Circuits (fig 1.6) by MJS Smith - permission requested)

There are many variants of the basic gate array including a channelled array (where the rows of cells are separated by a channel essentially to make the interconnect easier), a structured gate array (which combines some aspects of CBIC and the gate array).

In essence gate arrays remove more of the design process from the ASIC designer, making the process simpler and quicker at a loss of further flexibility. Hence complex designs are possible in gate arrays, which are economic at lower volumes still (typically 20k units/ year or lower).

For more information on gate arrays read sections 1.13 – 1.1.6 in the textbook.

1.1.5 Programmable logic devices (PLDs)

PLDs have been around for quite some time. They are based on ROM or PROM technology, are limited to a small number of gate equivalents, typically 1000, and are very easy to programme using PLD software such as PALASM or CUPL. Essentially all PLDs consist of an array of OR gates and an array of AND gates with some macrocells which are usually latches of flip-flops. Depending on the type one or other or both arrays are user programmable.

PLDs are very simple to use and reduce time/costs to very low levels making a PLD solution economic at 100 gates or lower.

For more information please read section 1.1.7 in the textbook.

1.1.5 Field Programmable Gate Arrays ( FPGA)

Making use of the increasing transistor densities and the non volatile nature of flash memory the field programmable gate array uses complex multi layer interconnects controlled by non volatile EEROM to provide interconnects for blocks containing multiple logical elements. The cost and density of these has dropped dramatically providing a quick and reliable route to prototyping and in manfacturing complex digital systems where the cost of an ASIC is unsupportable. The ability to rapidly can a digital design has made these invaluable in development and debugging applications.

[back to top]

1.2 ASIC Design Flow

The sequence of steps in designing an ASIC (design flow) is shown in Figure 1.3. The steps will now be listed and a brief description given of each.

Figure 1.3: Sequence of steps in designing an ASIC (design flow)

(from Application-Specific Integrated Circuits (fig 1.10) by MJS Smith - permission requested)

1. Design entry
• Using either schematic entry or more commonly now using a Hardware Description Language such as VHDL, Verilog, SystemVerilog.
  
  
2. Logic synthesis
• Use an HDL and a logic synthesis tool to produce a netlist, which is a description of the components and their connections.
• Significants amounts of information are required to provide the timing information for each block and for critical signals. Without these any synthesis is likely to fail. With modern DSM designs the impact of layout delays are far greater than gate ones and as such a synthesis tool such as Cadence PKS or Magma will handle synthesis, place and route, and timing closure.
  
3. System partitioning
• Divide the system into sensible blocks
• Modern tools tend to minimse the number of blocks on a system looking to merge structures built up as part of the HDL description of the system to increase the freedom the PKS tools have to accomodate timing requirments.
  
  
4. Simulation
• Check the system for functionality
• Check the system for timing closure to ensure it meets timing requirements.
  
  
5. Floorplanning
• Arrange the blocks on the chip
  
  
6. Placement
• Decide on locations of cells in a block.
  
  
7. Routing
• Make the connections between cells and blocks.
  
  
8. Circuit extraction
• Determine the resistances and capacitances of the interconnect
  
  
9. Post layout simulation
• Check the functionality again with the Rs and Cs of the interconnect included.

Modern synthesis tools will have physical knowledge of the chip layout and characteristics and will integrated the placement/routing/circuit extraction and Post Layout simulation stages automatically.

[back to top]

1.3 Case Study

[back to top]

1.4 ASIC Cost Comparisons

The economics of using ASICs in a product is an important consideration. Different ASIC solutions have different fixed costs and variable costs. To make realistic cost comparisons, costs must be up-to-date as they change often. In the example we use 'typical' costs to illustrate the differences.

• Read sections 1.4, 1.4.1, 1.4.2 in the textbook relating to product costs for different ASIC solutions.
  
• Figure 1.11 shows a break-even analysis for different ASIC types.

Figure 1.11: A break-even analysis for an FPGA, a masked gate array (MGA) and a custom cell-based ASIC (CBIC). The break-even volume between two technologies is the point at which the total cost of parts are equal. These numbers are very approximate.

(from Application-Specific Integrated Circuits (fig 1.11) by MJS Smith - permission requested)

• Familiarise yourself with the constituent parts of fixed costs (spreadsheet figure 1.12) - section 1.4.3 and variable costs (spreadsheet figure 1.14) - section 1.4.4 in the textbook.
  
• Understand how the costs change with technology advances and product maturity, see figure 1.15

Figure 1.15: Example price per gate figures.

(from Application-Specific Integrated Circuits (fig 1.15) by MJS Smith - permission requested)

[back to top]

1.5 Cell Libraries

• Read section 1.5 in the textbook.
• Understand the importance of the cell libraries in ASIC design.
• Be familiar with the different types of cell libraries  eg. physical, behavioural etc. and their uses in the design process.

[back to top]

1.6 Websites of Interest

[back to top]

1.7 CMOS Inverter

• Based on CMOS transistors. CMOS Complementary Metal Oxide Semiconductor, although the Metal oxide has long since been replaced with silicon dioxide.
• A CMOS transistor is essentially a switch with four terminals, gate(G), source(S), drain(D) and substrate or bulk (B).
• Two types of CMOS transistors  -  n-channel and p-channel.
• Positive logic assumed - VDD is logic 1 (say +5V) and VSS is logic 0 (say OV).
• n-channel transistor is ON with logic 1 at the gate and OFF with logic 0 at the gate.
• p-channel transistor is ON with logic 0 at the gate and OFF with logic 1 at the gate.  - see figure 2.1

Figure 2.1: CMOS transistors as switches

(a) An n-channel transistor. (b) a p-channel transistor. (c) A CMOS inverter and its symbol (an equilateral triangle and a circle).

(from Application-Specific Integrated Circuits (fig 2.1)
by MJS Smith - permission requested)

• CMOS inverter formed by connecting n-channel and p-channel transistors in series between VDD and VSS
• Operation is that inverter output at logic 0 for logic 1 input and output at logic 1 for logic 0 input - see CMOS interactive exercise.
• In both logic states one transistor OFF and hence no current flow and very low power dissipation (virtually zero) in both states.  Major advantage of CMOS!
• Other gates eg.  NAND/NOR and more complex structures can be easily designed - see figure 2.2 in the textbook.
• Theory of CMOS transistor operation complex - see sections 2.1, 2.1.1, 2.1.2, 2.1.4 in the textbook.
• Theory gives V-I equations in saturation (V_DS>V_GS-V_tN) as equation 2.12 and in linear region (V_DS < V_GS-V_tN) as equation 2.9 for n-channel transistor.
• Equations 2.15 apply for p-channel transistor.
• All V-I equations contain b term (Gain Factor) which equals k (process transconductance) times width/length of the transistor.
• b Important allowing current in a CMOS transistor to be varied by varying device geometry (W/L) as well as terminal voltages.
• Theory and practical measurements agree well (see figure 1.4). Short-channel transistors are normal in most ASIC devices.

Figure 2.4: MOS n-channel transistor characteristics for a generic 0.5mm process (G5).

A short channel transistor, with W = 6 mm and L = 0.6 mm (drawn) and a long channel transistor (W = 60mm, L = 6mm)

(from Application-Specific Integrated Circuits (fig 2.4)
by MJS Smith - permission requested)

• CAD programme SPICE often used to characterise transistors or gates.  Parameters for a generic 0.5µm CMOS process given in Table 2.1 for example.
• Due to transistor operation logic levels can be either strong or weak.
• n-channel transistor gives a strong '0' logic level but a weak '1' - see section 2.1.4 and Figure 2.5 in the textbook.
• p-channel transistor gives a strong '1' logic level but a weak '0'.
• Use both transistors together in CMOS gates to give strong '0' and '1' levels.

[back to top]

1.8  CMOS Process

• See section 2.2 in the textbook
• CMOS is the most common fabrication process for ASICs. It is illustrated in Figure 2.6 in the text book.
• The various mask/layer names together with MOSIS (US design house) mask labels are given in Table 2.2
• Using these names Figure 2.7 in the textbook shows the layers required to achieve a typical standard cell layout given in figure 1.3 (p8), together with the complete cell layout and the phantom layout often used in ASIC designs.
• 'Wells' of the opposite type of semiconductor to the substrate are used in CMOS to allow the fabrication of p-type and n-type transistors on the same substrate.
• Single, twin and triple-well processes available.
• In general the more wells in a process the more control over transistor properties.
• In all cases n-wells must be connected to the most positive part of the circuit (VDD) to ensure that substrate/source drain junctions are not forward biased and p-wells must be connected to the most negative part of the circuit (VSS) for the same reason.
• Often substrate connections not shown on circuit schematics but vital for correct circuit operation.
• CMOS process for circuit depicted in figure 2.7 described in pages 52-55 of the textbook.
• Sheet resistance is inversely proportional to the concentration of a doped layer. Sheet resistance is a measure of the concentration of a semiconductor doped layer.
• Sheet resistance measured in ohms/square since layers are very shallow compared to widths/lengths.
• Typical values between 1.1kW/sq for an n-well to 30 x 10^-3W/sq for metal - see Tables 2.3 and 2.4 in the textbook for example set.
• Contact resistance (CR) - metal/silicon - often significant and process steps taken to reduce CR and also improve contact reliability (see Tables 2.5 and 2.6 in the textbook).

[back to top]

1.9  CMOS Design Rules

• Design rules are a comprehensive set of rules, derived by a foundry which states the minimum geometry of layers and their relations to other layers to ensure correct circuit functionality after fabrication by the foundry.
• Often given in terms of λ (minimum feature size) currently 0.13µm so that the rules are scaleable as technology develops and feature size reduces.
• MOSIS is a US-based foundry serving their academic institutions; its design rules (version 7) are shown in figure 2.11, Table 2.7, 2.8 and 2.9 in the textbook.
• Design rules are only of concern to full-custom ASIC designs as in other ASICs the design rules are in-built into the layout and checking software and therefore it is impossible for designers to contravene them.

[back to top]

1.10 Logic Cells

1.10.1 Combinational

• Basic combinational gates (NAND, NOR, etc) can be made in CMOS (figure 2.2 in the textbook) as already discussed.
• More complex combinational cells comprising several gate combinations such as AND-OR-INVERT(AOI) and OR-AND-INVERT (OAI) are much more efficient in CMOS and often used in combinational design - see Table 2.10 in the textbook.
• Numbering notation based on number of inputs at first level and second level often used - see Figure 2.12 in the textbook.
• Design procedure (pushing bubbles!) using networks of transistors (stacks) used.
• Illustrated for the AOI 221in figure 2.13 of the textbook.
• Different hole and election mobilities give rise to different transistor gain factors bn and bp (equations 2.11 and 2.15 in the textbook).
• Equalise by adjusting (W/L) ratio of n and p type transistors to make bn and bp equal (same drive strength). (see section 2.4.2 in the textbook)
• Cells in a library available in a range of drive strengths.
• For transistors in series or parallel, design procedures are more complicated but essentially:
  
  
  • for transistors in parallel, make all the lengths unity and add the widths
  • for transistors in series, make all the widths unity and add the lengths.
    
    
• For example applied to AOI221 in figure 2.13c in the textbook.
• Alternative combinational design approach based on CMOS transmission gate (TX)  exists for simple gates. (section 2.4.3 and figure 2.14 in the textbook).
• More efficient in terms of number of transistors but other considerations may be important ie. charge sharing which may require extra buffering and therefore extra transistors.
• Can design of a 2/1 multiplexer using TX gates (figure 2.15 in the textbook)
• Comparison with design using OAI22 cell (figure 2.16 in the textbook) shows little difference but for longer MUXs differences can become significant.
• Can then design EXC-OR cell from a 2/1 MUX and an OR gate (section 2.4.4 in the textbook).

1.10.2 Sequential

• Synchronous design using a single system clock is nearly always used since it is safer, compatible with CAD tools and usually guarantees that the ASIC will work as simulated. Asynchronous designs are becoming popular but are more difficult.
• Sequential cells have a memory or storage feature
• Simple Latches are transparent (ie. changes at inputs appear at the Q output when the clock is high)
• Flip-flops are more complex (require at least two latches)
• Design a latch from TX gates - operation illustrated in Figure 2.17 in the textbook.
• Design a flip-flop from two D latches (see fig 2.18 in the textbook)
• Clocked inverter easily designed from an inverter and one TX gate (see fig 2.19 in the textbook) which can then replace inverters in latches and FFs

1.10.3 Datapath Logic Cells

• Datapath structures exploit the regularity of functions such as adders, multipliers, subtractors, barrel shifters etc. to give efficient VLSI implementations. For example, a 1-bit full adder output is usually expressed as:

              Sum  = A  B  C_IN
    and C_out   = A_.B + A_.C_IN + B_.C_IN

• These can be expressed in terms of the PARITY function (where an output is 1 if there are an odd number of inputs are 1's) and the MAJORITY function (where the output is 1 if the majority of the three inputs are 1) as:

           Sum  =  PARITY (A,B,C_IN)
    and C_out  =  MAJORITY (A,B,C_IN).

• Hence using form efficient single logic cells shown in figure 2.20a in the textbook.
• Layout in figure 2.20c with data running horizontally and control signals vertically and array structure shown in figure 2.20d, all in the textbook.
• Extend easily to a 4bit full adder (ADD4) by connecting four full adder cells as shown in figure 2.20b in the textbook.
• Datapath refers to the layout of buswide logic operating as in the ADD element described earlier (datapath cell or element).
• Datapath cells are usually stored in a library and are the same size so more complex datapath cells can be created (expandable).
• Usually orientated so that increasing size in bits grows the datapath in height; adding different elements to increase the function grows the datapath in width.
• Datapath implementations are regular and interconnect included in the cells.
• Disadvantages are control overheads and the requirement to pre-design the datapath cells themselves.

1.10.4 Datapath Cells

1.10.4.1 Adder

• Symbols given in figure 1.21

Figure 1.21: Symbols for a datapath adder.

(a) a data bus is shown by a heavy line and a bus symbol. If the bus is N-bits then MSB = n-1. (b) An alternative symbol for an adder. (c) Control signals are shown as lightweight lines.

• Table 2.11 reviews the common binary arithmetic operations (add, subtract etc) for the four common binary number representations (unsigned, signed magnitude, ones' complement, two's complement).
• Often addition is in terms of generate G(i) and propagate P(i) signals - see section 2.6.2 in the textbook for an explanation and equivalences.
• Form a ripple carry adder (RCA) conventionally (figure 1.22a) or using the generate/propagate approach (figure 1.22b).

Figure 1.22: The Ripple Carry Adder (RCA).

(a) A conventional RCA. The delay may be reduced slightly by adding pairs of bubbles as shown to use two-input NAND gates.

(b) An alternative RCA circuit topology using different cells for odd and even stages and an extra connection between cells.

The carry chain is a fast string of NAND gates (shown in bold).

• Other faster adders are based on carry-save (CSA) (figure 2.23 in the textbook), carry-by pass (CBA), carry-skip and the most well-known the carry-lookahead(CLA).
• Brent-Kung adder reduces the delay and increases the regularity of CLAs (figure 2.24 in the textbook).
• Fastest adders are based on carry-select leading to the conditional sum adder (CSA) (see figure 2.25 in the textbook).
• Graphs of normalized delays versus number of bits (figure 2.26a in the textbook) show ripple-carry to be the slowest, carry- select faster and carry-save the fastest.

• Graphs of areas versus bits (figure 2.26b in the textbook) show that ripple-carry and carry-save take up about the same area whereas carry-select takes up about twice as much area.

1.10.4.2 Multipliers

• Multiplication is a series of additions and shifts.
• Use a number of adders to form an array multiplier - see figure 2.27 in the textbook for a 6-bit multiplication illustration.
• Performance determined by the number of partial products and the addition of partial products.
• Use canonical signed - digit vectors (CSDs) to reduce the number of add/subtract operations and replace some additions by shifts.
• Further improvement by using Booth encoding (partial products reduced by a factor of 2 improves speed and area utilisation).
• Improve speed still further by using Wallace-trees and Dadda multipliers (Figures 2.28, 2.29 and 2.30 in the textbook)
• Several considerations apply in the choice of parallel multiplier architecture eg. overall speed, power dissipation, implementation (cell or full custom), pipelining etc.

1.10.4.3 Other datapath operations

• Combinational and sequential datapath cells eg.  NAND/NOR and FFS/latches available and operate identically to standard forms.
• Subtractors and adder/subtractors essentially adders with modified control lines.
• Barrel shifters rotate or shift input data stream by a specified amount - used in floating point arithmetic.
• Other floating point arithmetic operators include leading - one detectors, priority encoders, accumulators and registers of various types.

1.10.5 I/O Cells

• Popular tri-state bi-directional output buffer available (Figure 1.33)

Figure 1.33: Popular tri-state bi-directional output buffer

• Many other buffers available and easily designed
• I/O cells have to be designed to withstand static electricity electrostatic discharge (ESD) brought about by handling. Require input pads to be tied to structures that clamp the input voltage to below the gate breakdown voltage (10v for a 100Å gate oxide)

1.10.6 Compiled Cells

• Essentially automatic layout tools which generate regular structures of variable sizes such as RAM, ROM, multipliers.
• Often linked to a model compiler (for behavioural verification) and a netlist compiler (for layout level verification).
• Complete system produces blocks that are correct by design.

[back to top]

Partially Updated 30.08.11 NJC

Site Search

Powered by Google
Site Map