Performance Benchmarks
Click the button below to run all optimization benchmarks and compare performance.
Benchmark Results
| Test Case | Type | Baseline (ms) | Optimized (ms) | Speedup | Status |
|---|
Summary
Initializing benchmarks...
Base Converter
Convert numbers between different bases using optimized algorithms.
📝 How to enter numbers:
- Bases 2-10: Use digits 0-9 (e.g.,
1010for binary) - Bases 11-36: Use 0-9, A-Z (e.g.,
FFfor hex,Zfor base-36) - Bases 37-62: Use 0-9, a-z, A-Z (e.g.,
aZfor mixed case) - With prefixes:
xFF,0b1010,0o777for hex/binary/octal
Input
Base 11+: Use A-Z (10=A, 11=B, 35=Z) • Base 37+: Use a-z, A-Z
Base 10: Use digits 0-9
Output
Base 16: Use 0-9, A-F (e.g., FF, 1A2B)
-
-
Algorithm Used
Strategy:
-
Conversion Time:
-
Digits Processed:
-
Quick Presets
Number Representation Visualizer
Visualize how numbers are represented in different formats (integers, floating-point, Q-format).
📝 Abbreviation Explanation:
- S: Sign
- E: Exponent
- M: Mantissa
- I: Integer
- F: Fractional
Input Value
Enter a number (integer or decimal)
Representations
Integer Formats
INT4 (Signed)
Range: -8 to 7
-
-
INT8 (Signed)
Range: -128 to 127
-
-
INT16 (Signed)
Range: -32768 to 32767
-
-
INT32 (Signed)
Range: ±2.1B
-
-
INT64 (Signed)
Range: ±9.2E18
-
-
UINT8 (Unsigned)
Range: 0 to 255
-
-
Floating-Point Formats
FP16 (Half)
1s + 5e + 10m
-
-
-
FP32 (Single)
1s + 8e + 23m
-
-
-
FP64 (Double)
1s + 11e + 52m
-
-
-
BF16 (BFloat16)
1s + 8e + 7m
-
-
-
E2M1 (MXFP4)
1s + 2e + 1m
-
-
-
E8M0 (MXFP8)
1s + 8e + 0m*
-
-
-
E4M3 (MXFP8)
1s + 4e + 3m
-
-
-
E5M2 (MXFP8)
1s + 5e + 2m
-
-
-
E2M3 (MXFP6)
1s + 2e + 3m
-
-
-
E3M2 (MXFP6)
1s + 3e + 2m
-
-
-
Q Format (Fixed-Point)
Q7
1s + 7f
-
-
-
Q15
1s + 15f
-
-
-
Q31
1s + 31f
-
-
-
Q8.8
1s + 7i + 8f
-
-
-
Q16.16
1s + 15i + 16f
-
-
-
Q4.12
1s + 3i + 12f
-
-
-
Hexadecimal Representations
8-bit Hex
-
16-bit Hex
-
32-bit Hex
-
64-bit Hex
-
Results
Performance
Baseline Time
-
Optimized Time
-
Calculating...
Details
Test Type
-
Input Size
-
Status
-
0 ms
Baseline
0 ms
Optimized