cse477 lecture10

  • Upload
    ar

  • View
    224

  • Download
    0

Embed Size (px)

Citation preview

  • 8/8/2019 cse477 lecture10

    1/23

    CSE477 L10 Inverter, Dynamic.1 Irwin&Vijay, PSU, 2002

    CSE477VLSI Digital Circuits

    Fall 2002

    Lecture 10: The Inverter, A Dynamic View

    Mary Jane Irwin ( www.cse.psu.edu/~mji )www.cse.psu.edu/~cg477

    [Adapted from Rabaeys Digital Integrated Circuits, 2002, J. Rabaey et al.]

  • 8/8/2019 cse477 lecture10

    2/23

    CSE477 L10 Inverter, Dynamic.2 Irwin&Vijay, PSU, 2002

    Inverter Propagation Delay

    Propagation delay is proportional to the time-constant of

    the network formed by the pull-down resistor and the loadcapacitance

    tpHL = ln(2) Reqn CL = 0.69 Reqn CL

    tpLH = ln(2) Reqp CL = 0.69 Reqp CL

    tp = (tpHL + tpLH)/2 = 0.69 CL(Reqn + Reqp)/2

    To equalize rise and fall times make the on-resistance ofthe NMOS and PMOS approximately equal.

    VDD

    Rn

    Vout = 0

    Vin = V DD

    CL

    tpHL = f(Rn, CL)

  • 8/8/2019 cse477 lecture10

    3/23

    CSE477 L10 Inverter, Dynamic.3 Irwin&Vijay, PSU, 2002

    Inverter Transient Response

    -0.5

    0

    0.5

    1

    1.5

    2

    2.5

    3

    0 0.5 1 1.5 2 2.5

    Vin

    t (sec)x 10-10

    VDD=2.5V

    0.25QmW/Ln = 1.5

    W/Lp = 4.5

    Reqn= 13 k; (z 1.5)

    Reqp= 31 k; (z 4.5)

  • 8/8/2019 cse477 lecture10

    4/23

    CSE477 L10 Inverter, Dynamic.4 Irwin&Vijay, PSU, 2002

    Inverter Transient Response

    -0.5

    0

    0.5

    1

    1.5

    2

    2.5

    0 0.5 1 1.5 2 2.5

    Vin

    t (sec)x 10-10

    VDD=2.5V

    0.25QmW/Ln = 1.5

    W/Lp = 4.5

    Reqn= 13 k; (z 1.5)

    Reqp= 31 k; (z 4.5)

    tpHL = 36 psec

    tpLH = 29 psec

    so

    tp = 32.5 psec

    tf trtpHL tpLH

    From simulation: tpHL = 39.9 psec and tpLH = 31.7 psec

  • 8/8/2019 cse477 lecture10

    5/23

    CSE477 L10 Inverter, Dynamic. Irwin&Vijay, PSU, 2002

    Inverter Propagation Delay, Revisited

    To see how a designercan optimize the delay of a gate

    have to expand the Req in the delay equation

    1

    1.5

    2

    2.5

    3

    3.5

    4

    4.5

    5

    5.5

    0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4

    VDD (V)

    tpHL = 0.69 Reqn CL

    = 0.69 (3/4 (CL VDD)/IDSATn )

    } 0.52 CL / (W/Ln kn VDSATn )

  • 8/8/2019 cse477 lecture10

    6/23

    CSE477 L10 Inverter, Dynamic.6 Irwin&Vijay, PSU, 2002

    Design for Performance

    Reduce CL internal diffusion capacitance of the gate itself

    - keep the drain diffusion as small as possible

    interconnect capacitance

    fanout

    Increase W/L ratio of the transistor the most powerful and effective performance optimization

    tool in the hands of the designer

    watch out forself-loading! when the intrinsic capacitancedominates the extrinsic load

    Increase VDD can trade-off energy for performance

    increasing VDD above a certain level yields only very minimalimprovements

    reliability concerns enforce a firm upper bound on VDD

  • 8/8/2019 cse477 lecture10

    7/23

  • 8/8/2019 cse477 lecture10

    8/23

    CSE477 L10 Inverter, Dynamic.8 Irwin&Vijay, PSU, 2002

    PMOS/NMOS Ratio Effects

    .

    4

    4.

    1 2 4

    F = (W/Lp)/(W/Ln)

    x 10-11

    F of 2.4 (= 31 k;/13 k;)gives symmetrical

    response

    F of 1.6 to 1.9 givesoptimal performance

    tpLH

    tp

    tpHL

  • 8/8/2019 cse477 lecture10

    9/23

    CSE477 L10 Inverter, Dynamic.9 Irwin&Vijay, PSU, 2002

    Device Sizing for Performance

    Divide capacitive load,CL, into

    Cint : intrinsic - diffusion and Miller effect Cext : extrinsic - wiring and fanout

    tp = 0.69 Req Cint (1 + Cext/Cint) = tp0 (1 + Cext/Cint)

    where tp0 = 0.69 Req Cint is the intrinsic (unloaded) delay of thegate

    Widening both PMOS and NMOS by a factorS reducesReq by an identical factor (Req = Rref/S), but raises theintrinsic capacitance by the same factor (Cint = SCiref)

    tp

    = 0.69 Rref

    Ciref

    (1 + Cext

    /(SCiref

    )) = tp0

    (1 + Cext

    /(SCiref

    ))

    tp0 is independent of the sizing of the gate; with no load the driveof the gate is totally offset by the increased capacitance

    any S sufficiently larger than (Cext/Cint) yields the bestperformance gains with least area impact

  • 8/8/2019 cse477 lecture10

    10/23

    CSE477 L10 Inverter, Dynamic.10 Irwin&Vijay, PSU, 2002

    Sizing Impacts on Delay

    2

    2.2

    2.4

    2.

    2.

    .2

    .4

    .

    .

    1 7 11 1 1

    S

    x 10-11 The majority of theimprovement is already

    obtained for S = 5. Sizing

    factors larger than 10

    barely yield any extra gain

    (and cost significantlymore area).

    for a fixed load

    self-loading effect

    (intrinsic capacitance

    dominates)

  • 8/8/2019 cse477 lecture10

    11/23

    CSE477 L10 Inverter, Dynamic.11 Irwin&Vijay, PSU, 2002

    Impact of Fanout on Delay

    Extrinsic capacitance, Cext, is a function of the fanout of

    the gate - the larger the fanout, the larger the externalload.

    First determine the input loading effect of the inverter.

    Both Cg and Cint are proportional to the gate sizing, soCint = KCg is independent of gate sizing and

    tp = tp0 (1 + Cext/ KCg) = tp0 (1 + f/K)

    i.e., the delay of an inverter is a function of the ratio

    between its external load capacitance and its input gatecapacitance: the effective fan-out f

    f = Cext/Cg

  • 8/8/2019 cse477 lecture10

    12/23

    CSE477 L10 Inverter, Dynamic.12 Irwin&Vijay, PSU, 2002

    Inverter Chain

    If CL is given

    How should the inverters be sized?

    How many stages are needed to minimize the delay?

    In Out

    CL

    Real goal is to minimize the delay through an inverter

    chain

    the delay of the j-th inverter stage is

    tp,j = tp0 (1 + Cg,j+1/(KCg,j)) = tp0(1 + fj/ K)

    and tp = tp1 + tp2 + . . . + tpN

    so tp = tp,j = tp0 (1 + Cg,j+1/(KCg,j))

    Cg,11 2 N

  • 8/8/2019 cse477 lecture10

    13/23

    CSE477 L10 Inverter, Dynamic.13 Irwin&Vijay, PSU, 2002

    Sizing the Inverters in the Chain

    The optimum size of each inverter is the geometric mean

    of its neighbors meaning that if each inverter is sized upby the same factor f wrt the preceding gate, it will have thesame effective fan-out and the same delay

    f = CL/Cg,1 = F

    where F represents the overall effective fan-out of thecircuit (F = CL/Cg,1)

    and the minimum delay through the inverter chain is

    tp = N tp0 (1 + ( F ) / K)

    The relationship between tp and F is linear for one inverter,square root for two, etc.

    N N

    N

  • 8/8/2019 cse477 lecture10

    14/23

    CSE477 L10 Inverter, Dynamic.14 Irwin&Vijay, PSU, 2002

    Example of Inverter Chain Sizing

    CL/Cg,1 has to be evenly distributed over N = 3 inverters

    CL/Cg,1 = 8/1f =

    In Out

    CL = 8 Cg,1Cg,11

  • 8/8/2019 cse477 lecture10

    15/23

    CSE477 L10 Inverter, Dynamic.1 Irwin&Vijay, PSU, 2002

    Example of Inverter Chain Sizing

    CL/Cg,1 has to be evenly distributed over N = 3 inverters

    CL/Cg,1 = 8/1f =

    In Out

    CL = 8 Cg,1Cg,11 f = 2 f 2 = 4

    38 = 2

  • 8/8/2019 cse477 lecture10

    16/23

    CSE477 L10 Inverter, Dynamic.16 Irwin&Vijay, PSU, 2002

    Determining N: Optimal Number of Inverters

    What is the optimal value for N given F (=fN) ?

    if the number of stages is too large, the intrinsic delay of thestages becomes dominate

    if the number of stages is too small, the effective fan-out of eachstage becomes dominate

    N N

    The optimum N is found by differentiating the minimumdelay expression divided by the number of stages andsetting the result to 0, giving

    K + F - ( F lnF)/N = 0

    ForK = 0 (ignoring self-loading) N = ln (F) and theeffective-fan out becomes f = e = 2.71828

    ForK = 1 (the typical case) the optimum effective fan-out(tapering factor) turns out to be close to 3.6

  • 8/8/2019 cse477 lecture10

    17/23

    CSE477 L10 Inverter, Dynamic.17 Irwin&Vijay, PSU, 2002

    Optimum Effective Fan Out

    Choosing f larger than optimum has little effect on delayand reduces the number of stages (and area).

    Common practice to use f = 4 (forK = 1)

    But too many stages has a substantial negative impact on delay

    2.5

    3

    3.5

    4

    4.5

    5

    0 0.5 1 1.5 2 2.5 3

    K

    0

    1

    2

    3

    4

    5

    6

    7

    1 1.5 2 2.5 3 3.5 4 4.5 5

    f

  • 8/8/2019 cse477 lecture10

    18/23

    CSE477 L10 Inverter, Dynamic.18 Irwin&Vijay, PSU, 2002

    Example of Inverter (Buffer) Staging

    CL = 64 Cg,1Cg,1 = 1

    1

    CL = 64 Cg,1Cg,1 = 1

    1 8

    CL = 64 Cg,1Cg,1 = 1

    1 4 16

    CL = 64 Cg,1Cg,1 = 1

    1 2.8 8 22.6

    N f tp

    1 64 65

    2 8 18

    3 4 15

    4 2.8 15.3

  • 8/8/2019 cse477 lecture10

    19/23

  • 8/8/2019 cse477 lecture10

    20/23

    CSE477 L10 Inverter, Dynamic.20 Irwin&Vijay, PSU, 2002

    Input Signal Rise/Fall Time

    In reality, the input signal

    changes gradually (and bothPMOS and NMOS conduct fora brief time). This affects thecurrent available forcharging/discharging CL and

    impacts propagation delay.

    .

    .

    .

    .

    .

    .

    .

    .

    ts(sec)

    x 1 -11

    x 1 -11

    for a minimum-size inverter

    with a fan-out of a single gate

    tp increases linearly withincreasing input slope, ts,once ts > tp

    ts is due to the limited drivingcapability of the preceding gate

  • 8/8/2019 cse477 lecture10

    21/23

    CSE477 L10 Inverter, Dynamic.21 Irwin&Vijay, PSU, 2002

    Design Challenge

    A gate is never designed in isolation: its performance isaffected by both the fan-out and the driving strength of thegate(s) feeding its inputs.

    tip = tistep + L t

    i-1step (L } 0.25)

    Keep signal rise times smaller than or equal to the gatepropagation delays.

    good for performance

    good for power consumption

    Keeping rise and fall times of the signals small and ofapproximately equal values is one of the major challengesin high-performance designs - slope engineering.

  • 8/8/2019 cse477 lecture10

    22/23

    CSE477 L10 Inverter, Dynamic.22 Irwin&Vijay, PSU, 2002

    Delay with Long Interconnects

    When gates are farther apart, wire capacitance and

    resistance can no longer be ignored.

    tp = 0.69RdrCint + (0.69Rdr+0.38Rw)Cw + 0.69(Rdr+Rw)Cfan

    where Rdr= (Reqn + Reqp)/2

    = 0.69Rdr(Cint+Cfan) + 0.69(Rdrcw+rwCfan)L + 0.38rwcwL2

    cint

    Vin

    cfan

    (rw, cw, L) Vout

    Wire delay rapidly becomes the dominate factor (due tothe quadratic term) in the delay budget for longer wires.

  • 8/8/2019 cse477 lecture10

    23/23

    CSE477 L10 Inverter, Dynamic.23 Irwin&Vijay, PSU, 2002

    Next Lecture and Reminders

    Next lecture

    Designing fast logic- Reading assignment Rabaey, et al, 6.2.1

    Reminders

    Project specifications due today

    HW3 due next Thursday, Oct 10th

    (hand in to TA) Class cancelled on Oct 10th as make up for evening midterm

    I will be out of town Oct 10th through Oct 15th and Oct 18th

    through Oct 23rd, so office hours during those periods arecancelled

    We will have a guest lecturer on Oct 22

    nd

    Evening midterm exam scheduled

    - Wednesday, October 16th from 8:15 to 10:15pm in 260 Willard

    - Only one midterm conflict filed for so far