ESP32-S3 vs Feather M4
Xtensa LX7 (ESP32-S3) at 240 MHz against Cortex-M4F (SAMD51) at 120 MHz, on real hardware, same C kernels, clocks verified.
ESP32-S3
Xtensa LX7 (ESP32-S3)
- Family
- Xtensa
- Board
- ESP32-S3 DevKit
- Clock
- 240 MHz
- float / double
- 32-bit / 64-bit
Feather M4
Cortex-M4F (SAMD51)
- Family
- Arm Cortex-M
- Board
- Feather M4 Express
- Clock
- 120 MHz
- float / double
- 32-bit / 64-bit
Typical gap
1.48×
ESP32-S3 faster, geometric mean of 30 cells
Biggest win for ESP32-S3
2.87×
double modulo
Biggest win for Feather M4
2.44×
double divide
Clock for clock
1.35×
Feather M4 needs fewer cycles per op (2.00× clock gap)
Operation by operation
Every cell
How many times faster ESP32-S3 is than Feather M4, operations per second. Blue: ESP32-S3 ahead. Orange: Feather M4 ahead.
| Add | Subtract | Multiply | Divide | Modulo | |
|---|---|---|---|---|---|
| uint8_t | 1.41× 12.7 M/s · 8.98 M/s | 1.41× 12.7 M/s · 8.98 M/s | 1.41× 12.7 M/s · 8.98 M/s | 1.64× 11.5 M/s · 7.01 M/s | 1.72× 10.9 M/s · 6.36 M/s |
| uint16_t | 1.41× 12.7 M/s · 8.98 M/s | 1.41× 12.7 M/s · 8.98 M/s | 1.41× 12.7 M/s · 8.98 M/s | 1.66× 10.1 M/s · 6.12 M/s | 1.81× 9.72 M/s · 5.36 M/s |
| uint32_t | 1.38× 13.4 M/s · 9.71 M/s | 1.38× 13.4 M/s · 9.71 M/s | 1.38× 13.4 M/s · 9.71 M/s | 1.68× 11 M/s · 6.53 M/s | 1.69× 10.5 M/s · 6.19 M/s |
| uint64_t | 1.27× 7.87 M/s · 6.19 M/s | 1.26× 7.81 M/s · 6.19 M/s | 1/1.03× 5.45 M/s · 5.61 M/s | 2.03× 1.8 M/s · 885 k/s | 1.97× 1.74 M/s · 885 k/s |
| float | 1.41× 12.7 M/s · 8.98 M/s | 1.41× 12.7 M/s · 8.98 M/s | 1.41× 12.7 M/s · 8.98 M/s | 1/1.36× 3.48 M/s · 4.73 M/s | 2.25× 1.95 M/s · 866 k/s |
| double | 2.54× 3.04 M/s · 1.2 M/s | 2.34× 2.79 M/s · 1.19 M/s | 1.17× 2 M/s · 1.7 M/s | 1/2.44× 516 k/s · 1.26 M/s | 2.87× 1.22 M/s · 425 k/s |