Int vs Float in C++

Int vs Float in C++: Performance-Oriented Data Type Choice

---

Prerequisites

---

1. Why Data Type Choice Matters

Not all numeric types are equal in performance.

The type you choose affects:

• CPU execution speed
• Precision
• Cache efficiency
• Overall performance

If exact floating-point values are NOT required, prefer integer-based computation

Feature	int	float
Precision	Exact	Approximate
Speed	Faster (generally)	Slower (depends on hardware)
Determinism	High	Low (rounding errors)
Comparison	Reliable	Risky

2. Int vs Float

✔ Reasons of using Int

• Simpler CPU instructions
• No rounding / normalization
• Better predictability
• Often lower latency

Comparison Problem with Float

❌ Unsafe

if (a == b) 
{
    // may fail due to precision error
}

✔ Safer

if (fabs(a - b) < 1e-6) 
{
    // approximate comparison
}

Extra computation required → slower

3. Replace Float with Int (Scaling Technique)

Convert float to scaled integer

float value = 3.14f;
int scaled = (int)(value * 100);  // 314

Now compare as integer:

if (scaled > 300) 
{
    // fast integer comparison
}

❌ Float-based

if (distance < 1.5f) 
{
    // ...
}

✔ Int-based

int dist_scaled = distance * 100;
if (dist_scaled < 150) 
{
    // ...
}

Faster + deterministic

3-1. When to Prefer int

✔ Only comparison matters
✔ Fixed precision acceptable
✔ Performance-critical loops
✔ Large-scale data processing

3-2. When to Prefer float

✔ Real-world measurement
✔ Continuous values
✔ Scientific / graphics computation
✔ High precision required

Approach	Pros	Cons
int	Fast, deterministic	Limited range/precision
float	Flexible, expressive	Slower, precision issues