ESP32-C6 vs Feather M4
RV32IMAC (ESP32-C6) at 160 MHz against Cortex-M4F (SAMD51) at 120 MHz, on real hardware, same C kernels, clocks verified.
ESP32-C6
RV32IMAC (ESP32-C6)
- Family
- RISC-V
- Board
- ESP32-C6-DevKitM-1
- Clock
- 160 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
1.03×
Feather M4 faster, geometric mean of 30 cells
Biggest win for ESP32-C6
1.56×
uint8_t subtract
Biggest win for Feather M4
5.23×
float divide
Clock for clock
1.38×
Feather M4 needs fewer cycles per op (1.33× clock gap)
Operation by operation
Every cell
How many times faster ESP32-C6 is than Feather M4, operations per second. Blue: ESP32-C6 ahead. Orange: Feather M4 ahead.
| Add | Subtract | Multiply | Divide | Modulo | |
|---|---|---|---|---|---|
| uint8_t | 1.43× 12.8 M/s · 8.98 M/s | 1.56× 14 M/s · 8.98 M/s | 1.56× 14 M/s · 8.98 M/s | 1.15× 8.1 M/s · 7.01 M/s | 1.29× 8.2 M/s · 6.36 M/s |
| uint16_t | 1.15× 10.3 M/s · 8.98 M/s | 1.15× 10.3 M/s · 8.98 M/s | 1.23× 11.1 M/s · 8.98 M/s | 1.20× 7.35 M/s · 6.12 M/s | 1.39× 7.43 M/s · 5.36 M/s |
| uint32_t | 1.32× 12.8 M/s · 9.71 M/s | 1.22× 11.9 M/s · 9.71 M/s | 1.22× 11.9 M/s · 9.71 M/s | 1.12× 7.35 M/s · 6.53 M/s | 1.20× 7.43 M/s · 6.19 M/s |
| uint64_t | 1.32× 8.2 M/s · 6.19 M/s | 1.40× 8.65 M/s · 6.19 M/s | 1.26× 7.1 M/s · 5.61 M/s | 1.14× 1.01 M/s · 885 k/s | 1.11× 986 k/s · 885 k/s |
| float | 1/3.74× 2.4 M/s · 8.98 M/s | 1/3.92× 2.29 M/s · 8.98 M/s | 1/3.33× 2.7 M/s · 8.98 M/s | 1/5.23× 905 k/s · 4.73 M/s | 1.11× 959 k/s · 866 k/s |
| double | 1.17× 1.39 M/s · 1.2 M/s | 1.14× 1.35 M/s · 1.19 M/s | 1/1.00× 1.7 M/s · 1.7 M/s | 1/2.56× 492 k/s · 1.26 M/s | 1.52× 647 k/s · 425 k/s |