Arduino Uno vs Feather M4
ATmega328P at 16 MHz against Cortex-M4F (SAMD51) at 120 MHz, on real hardware, same C kernels, clocks verified.
Feather M4
Cortex-M4F (SAMD51)
- Family
- Arm Cortex-M
- Board
- Feather M4 Express
- Clock
- 120 MHz
- float / double
- 32-bit / 64-bit
Typical gap
41.8×
Feather M4 faster, geometric mean of 30 cells
Biggest win for Arduino Uno
0.24×
double modulo
Biggest win for Feather M4
263×
uint32_t divide
Clock for clock
5.58×
Feather M4 needs fewer cycles per op (7.50× clock gap)
Operation by operation
Every cell
How many times faster Arduino Uno is than Feather M4, operations per second. Blue: Arduino Uno ahead. Orange: Feather M4 ahead.
| Add | Subtract | Multiply | Divide | Modulo | |
|---|---|---|---|---|---|
| uint8_t | 1/18.3× 490 k/s · 8.98 M/s | 1/18.3× 490 k/s · 8.98 M/s | 1/20.0× 448 k/s · 8.98 M/s | 1/49.3× 142 k/s · 7.01 M/s | 1/44.7× 142 k/s · 6.36 M/s |
| uint16_t | 1/17.2× 522 k/s · 8.98 M/s | 1/17.2× 522 k/s · 8.98 M/s | 1/21.7× 413 k/s · 8.98 M/s | 1/88.0× 69.5 k/s · 6.12 M/s | 1/77.2× 69.5 k/s · 5.36 M/s |
| uint32_t | 1/28.3× 342 k/s · 9.71 M/s | 1/28.4× 342 k/s · 9.71 M/s | 1/71.9× 135 k/s · 9.71 M/s | 1/263× 24.8 k/s · 6.53 M/s | 1/250× 24.8 k/s · 6.19 M/s |
| uint64_t | 1/83.1× 74.6 k/s · 6.19 M/s | 1/83.1× 74.6 k/s · 6.19 M/s | 1/156× 35.9 k/s · 5.61 M/s | 1/57.9× 15.3 k/s · 885 k/s | 1/58.2× 15.2 k/s · 885 k/s |
| float | 1/91.4× 98.3 k/s · 8.98 M/s | 1/90.9× 98.7 k/s · 8.98 M/s | 1/103× 87 k/s · 8.98 M/s | 1/147× 32.2 k/s · 4.73 M/s | 1/8.43× 103 k/s · 866 k/s |
| double | 1/12.2× 98.1 k/s · 1.2 M/s | 1/12.1× 98.6 k/s · 1.19 M/s | 1/19.6× 86.8 k/s · 1.7 M/s | 1/39.1× 32.2 k/s · 1.26 M/s | 1/4.14× 103 k/s · 425 k/s |