module sum(c, pg, p, s);
    input c;
    input pg;
    input p;
    output s;
    assign s = c ? pg : p;
endmodule

module pg(x, y, g, p, pg);
    input x;
    input y;
    output g;
    output p;
    output pg;
    assign g = x & y;
    assign pg = y ? x : ~x;
    assign p = ~pg;
endmodule

module cla(p0, g0, p1, g1, p2, g2, p3, g3, c0, c1, c2, c3, Gg, Gp);
    input p0;
    input g0;
    input p1;
    input g1;
    input p2;
    input g2;
    input p3;
    input g3;
    input c0;
    output c1;
    output c2;
    output c3;
    output Gg;
    output Gp;
    assign c1 = g0 | p0 & c0;
    assign c2 = g1 | p1 & g0 | p1 & p0 & c0;
    assign c3 = g2 | p2 & g1 | p2 & p1 & g0 | p2 & p1 & p0 & c0;
    assign Gg = g3 | p3 & g2 | p3 & p2 & g1 | p3 & p2 & p1 & g0;
    assign Gp = p3 & p2 & p1 & p0;
endmodule

module clasum4(x, y, cin, s, p, g, cout);
    input [3:0] x;
    input [3:0] y;
    input cin;     
    output [3:0] s;
    output [3:0] p;
    output [3:0] g;
    output cout;
    wire c1, c2, c3;
 
    assign g[0] = x[0] & y[0], g[1] = x[1] & y[1], g[2] = x[2] & y[2], g[3] = x[3] & y[3]; 
    assign p[0] = x[0] ^ y[0], p[1] = x[1] ^ y[1], p[2] = x[2] ^ y[2], p[3] = x[3] ^ y[3];
 
    assign c1 = g[0] | (p[0] & cin);
    assign c2 = g[1] | (p[1] & g[0]) | (p[1] & p[0] & cin);
    assign c3 = g[2] | (p[2] & g[1]) | (p[2] & p[1]) | (p[2] & p[1] & p[0] & cin);
    assign cout = g[3] | (p[3] & g[2]) | (p[3] & p[2]) | (p[3] & p[2] & p[1]) | (p[3] & p[2] & p[1] & p[0] & cin);
 
    assign s[0] = cin ^ p[0], s[1] = c1 ^ p[1], s[2] = c2 ^ p[2], s[3] = c3 ^ p[3];
endmodule

Alternativní řešení 4bit sčítačky pomocí modulů viz výše:

module add4(x, y, c, s, Gp, Gg);
    input [3:0] x;
    input [3:0] y;
    input c;
    output [3:0] s;
    output Gp;
    output Gg;    
    wire p0, g0, pg0, p1, g1, pg1, p2, g2, pg2, p3, g3, pg3, c1, c2, c3;   
    sum sum0(c, pg0, p0, s[0]);
    sum sum1(c1, pg1, p1, s[1]);
    sum sum2(c2, pg2, p2, s[2]);
    sum sum3(c3, pg3, p3, s[3]);
    pg pg_0(x[0], y[0], g0, p0, pg0);
    pg pg_1(x[1], y[1], g1, p1, pg1);
    pg pg_2(x[2], y[2], g2, p2, pg2);
    pg pg_3(x[3], y[3], g3, p3, pg3);
    cla cla(p0, g0, p1, g1, p2, g2, p3, g3, c, c1, c2, c3, Gg, Gp);
endmodule

module add16(x, y, c, s, Sp, Sg);
    input [15:0] x;
    input [15:0] y;
    input c;
    output [15:0] s;
    output Sp;
    output Sg;
    wire Gp0, Gg0, Gp1, Gg1, Gp2, Gg2, Gp3, Gg3, c, Gc1, Gc2, Gc3;
    add4 add0_3(x[3:0], y[3:0], c, s[3:0], Gp0, Gg0);
    add4 add4_7(x[7:4], y[7:4], Gc1, s[7:4], Gp1, Gg1);
    add4 add8_11(x[11:8], y[11:8], Gc2, s[11:8], Gp2, Gg2);
    add4 add12_15(x[15:12], y[15:12], Gc3, s[15:12], Gp3, Gg3);
    cla gcla(Gp0, Gg0, Gp1, Gg1, Gp2, Gg2, Gp3, Gg3, c, Gc1, Gc2, Gc3, Sg, Sp);
endmodule

module add64(x, y, c, s, Np, Ng, cout);
    input [63:0] x;
    input [63:0] y;
    input c;
    output [63:0] s;
    output Np;
    output Ng;
    output cout;
    wire Sp0, Sg0, Sp1, Sg1, Sp2, Sg2, Sp3, Sg3, c, Sc1, Sc2, Sc3;
    add16 add0_15(x[15:0], y[15:0], c, s[15:0], Sp0, Sg0);
    add16 add16_31(x[31:16], y[31:16], Sc1, s[31:16], Sp1, Sg1);
    add16 add32_47(x[47:32], y[47:32], Sc2, s[47:32], Sp2, Sg2);
    add16 add48_63(x[63:48], y[63:48], Sc3, s[63:48], Sp3, Sg3);
    cla scla(Sp0, Sg0, Sp1, Sg1, Sp2, Sg2, Sp3, Sg3, c, Sc1, Sc2, Sc3, Ng, Np);
    assign cout = Ng | Np & c;
endmodule

Vyšlo mi:

• 3 úrovně CLA: 3(pg)+2*2(sejít na spodní CLA)+3*2(vystoupat přes c)+2(sum)=15
• 3(1. sečtení)+63*2(průchod c)=129
• zhruba 9x rychlejší (8,6 nebo 9,2, pokud se NOT nepočítá jako hradlo)