Python Big Number & Overhead Lab

Profiler Controls Interactive

Target Big Number Operation

Input Value (n = 5000)

Simulated Loop Iterations (Eval Loop) 100,000

Dynamic Type-Inspection Overhead

Simulates Python's requirement to inspect PyObject->ob_type and dispatch via tp_as_number->nb_add instead of direct CPU register addition.

Quick Representative Presets

Quora Core Finding: In Python, integers have arbitrary precision (handled via PyLongObject arrays of 30-bit limbs). The slowness comes from dynamic operand type checks on every + opcode and pointer chasing.

CPython Performance Telemetry Ready & Verified

Computed Digits

16,325

Exact decimal length

Execution Time

42.5 ms

Includes CPy dispatch

Memory Consumed

128.4 KB

PyLongObject limb array

Overhead Penalty

68.4%

Dynamic typing & dispatch

CPython 3.12 Disassembly (Eval Loop)

# dis.dis(calc_loop) 0 RESUME 0 2 LOAD_FAST 0 (result) 4 LOAD_FAST 1 (operand) 6 BINARY_OP 0 (+) ; <-- Dynamic dispatch traps here 10 STORE_FAST 0 (result) 12 JUMP_BACKWARD 6 (to 2)

Runtime Dispatch Pipeline (Per Single '+')

1. TYPE CHECK Inspect left->ob_type and right->ob_type (Are they strings, ints, floats, or custom?)

2. SLOT DISPATCH Traverse tp_as_number->nb_add pointer table; verify no string concatenation.

3. LIMB ARITHMETIC Execute arbitrary-precision Karatsuba / grade-school limb addition in C.

4. BOXING Allocate fresh heap PyLongObject container with reference count = 1.

Result Digits Sample First 30 ... Last 30 Digits

422857792640554387... [calculating digits] ... 000000000000000000