Decimal value for IEEE 754 = (-1)^S×(1+.M)×2^E-bias

S = 0 (since it is a positive number)

Exponent:
Exponent = ^log(4.00012)⁄_log(2)≈2
(E = exponent + bias = 2 + 127 = 129)
E = 1000 0001₂

1+.M = ^4.00012⁄_2²≈1.00003
.M ≈ 0.00003
Convert 0.00003 to binary:

Multiply by 2 for Fraction Part:
0.00003	× 2 =	0.00006		...0
0.00006	× 2 =	0.00012		...0
0.00012	× 2 =	0.00024		...0
0.00024	× 2 =	0.00048		...0
0.00048	× 2 =	0.00096		...0
0.00096	× 2 =	0.00192		...0
0.00192	× 2 =	0.00384		...0
0.00384	× 2 =	0.00768		...0
0.00768	× 2 =	0.01536		...0
0.01536	× 2 =	0.03072		...0
0.03072	× 2 =	0.06144		...0
0.06144	× 2 =	0.12288		...0
0.12288	× 2 =	0.24576		...0
0.24576	× 2 =	0.49152		...0
0.49152	× 2 =	0.98304		...0
0.98304	× 2 =	1.96608		...1
0.96608	× 2 =	1.93216		...1
0.93216	× 2 =	1.86432		...1
0.86432	× 2 =	1.72864		...1
0.72864	× 2 =	1.45728		...1
0.45728	× 2 =	0.91456		...0
0.91456	× 2 =	1.82912		...1
0.82912	× 2 =	1.65824		...1
0.65824	× 2 =	1.31648		...1

Fraction Binary: .000000000000000111110111₂
Fraction binary for single precision mantissa = .00000000000000011111100(23-bit length only) (Since the last 24 bit fraction is 1, we add 1 to the remaining 23-bit binary to round the number)
= .0000 0000 0000 0001 1111 100


Binary of to IEEE single-precision floating point number:

Sign	Exponent		Mantissa
0	10000001		0000 0000 0000 0001 1111 100
Hexadecimal digit sequence:
0100	0000	1000		0000	0000	0000	1111	1100
4	0	8		0	0	0	F	C

4.00012 to IEEE single-precision floating point number in hexadecimal digit sequence is 408000FCH.

Datum of address ADDR+00H to ADDR+03H in big endian:

Address	Content
ADDR+00H	40
ADDR+01H	80
ADDR+02H	00
ADDR+03H	FC