🔁 Sign Extension in 2's Complement

Sign extension is the process of increasing the bit width of a signed binary number by replicating its sign bit (most significant bit) into the new higher-order bits. This ensures the number retains its original value when represented in a wider format.

• For positive numbers (MSB = 0): pad with 0s
• For negative numbers (MSB = 1): pad with 1s

🍯 Value Preservation

Sign extension ensures that the numerical meaning of a signed binary number remains unchanged when its bit width increases.

• Example: 1101 (−3 in 4-bit 2’s complement)
  • Without sign extension → 00001101 (+13) ❌
  • With sign extension → 11111101 (−3) ✅

❓ Why does arithmetic break without sign extension?

Because you’re doing correct math on the wrong number. If you pad a negative number with zeros instead of replicating the sign bit, you change its meaning entirely.

Example: 4-bit: 1101 (−3) 8-bit: 00000101 (+5)

Without sign extension: 1101 → 00001101 (+13) 00001101 + 00000101 = 00010010 (+18) ❌

With sign extension: 1101 → 11111101 (−3) 11111101 + 00000101 = 00000010 (+2) ✅

❓ Why is 2’s complement arithmetic bitwise consistent only if sign extension is applied?

Because each bit has a fixed weight:

• LSB: $+2^0$
• …
• MSB: $-2^{n-1}$ ← this encodes the sign

Sign extension preserves this weight structure when increasing bit width. Without it, the MSB loses its negative weight, and the number is misinterpreted.

Hardware Simplicity

❓How does sign extension simplify hardware?

It allows the ALU to treat all operands as uniform bit patterns, regardless of their original size.

• ALUs operate on fixed-width inputs (e.g. 32-bit)
• Sign extension ensures smaller operands are padded without changing their value
• No need for:
  • Special logic to detect operand size
  • Conditional branching for signed vs unsigned
  • Manual sign interpretation

This enables:

• Reuse of logic gates across operand sizes
• Consistent carry chains and overflow detection
• Simplified control signals—sign is embedded in the bits

🧪 Behavior Summary

🔍 Bitwise Consistency Explained

2’s complement arithmetic is bitwise consistent across widths only if sign extension is applied.

• Each bit has a fixed weight; MSB has negative weight
• Sign extension ensures that this weight is preserved when increasing bit width
• Arithmetic operations (add, subtract) behave identically across widths because the sign is embedded in the bit pattern

Without sign extension:

• You change the sign and magnitude
• You break the consistency of 2’s complement logic

❓ Does this apply only to negative numbers?

Yes. Positive numbers already have leading zeros, so padding with more zeros doesn’t change their value. Negative numbers rely on the MSB’s negative weight, which must be preserved by replicating the 1s during extension.

🧼 Summary Table