ESP32-S3 vs ESP32-C6
Xtensa LX7 (ESP32-S3) at 240 MHz against RV32IMAC (ESP32-C6) at 160 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
ESP32-C6
RV32IMAC (ESP32-C6)
- Family
- RISC-V
- Board
- ESP32-C6-DevKitM-1
- Clock
- 160 MHz
- float / double
- 32-bit / 64-bit
Typical gap
1.53×
ESP32-S3 faster, geometric mean of 30 cells
Biggest win for ESP32-S3
5.54×
float subtract
Biggest win for ESP32-C6
1.30×
uint64_t multiply
Clock for clock
1.02×
ESP32-S3 needs fewer cycles per op (1.50× clock gap)
Operation by operation
Every cell
How many times faster ESP32-S3 is than ESP32-C6, operations per second. Blue: ESP32-S3 ahead. Orange: ESP32-C6 ahead.
| Add | Subtract | Multiply | Divide | Modulo | |
|---|---|---|---|---|---|
| uint8_t | 1/1.01× 12.7 M/s · 12.8 M/s | 1/1.10× 12.7 M/s · 14 M/s | 1/1.10× 12.7 M/s · 14 M/s | 1.42× 11.5 M/s · 8.1 M/s | 1.34× 10.9 M/s · 8.2 M/s |
| uint16_t | 1.23× 12.7 M/s · 10.3 M/s | 1.23× 12.7 M/s · 10.3 M/s | 1.15× 12.7 M/s · 11.1 M/s | 1.38× 10.1 M/s · 7.35 M/s | 1.31× 9.72 M/s · 7.43 M/s |
| uint32_t | 1.04× 13.4 M/s · 12.8 M/s | 1.13× 13.4 M/s · 11.9 M/s | 1.13× 13.4 M/s · 11.9 M/s | 1.49× 11 M/s · 7.35 M/s | 1.41× 10.5 M/s · 7.43 M/s |
| uint64_t | 1/1.04× 7.87 M/s · 8.2 M/s | 1/1.11× 7.81 M/s · 8.65 M/s | 1/1.30× 5.45 M/s · 7.1 M/s | 1.78× 1.8 M/s · 1.01 M/s | 1.77× 1.74 M/s · 986 k/s |
| float | 5.28× 12.7 M/s · 2.4 M/s | 5.54× 12.7 M/s · 2.29 M/s | 4.70× 12.7 M/s · 2.7 M/s | 3.84× 3.48 M/s · 905 k/s | 2.03× 1.95 M/s · 959 k/s |
| double | 2.18× 3.04 M/s · 1.39 M/s | 2.06× 2.79 M/s · 1.35 M/s | 1.17× 2 M/s · 1.7 M/s | 1.05× 516 k/s · 492 k/s | 1.89× 1.22 M/s · 647 k/s |