Thing Done

← The fleet

Feather M4 vs Pi Pico 2W (RISC-V)

Cortex-M4F (SAMD51) at 120 MHz against Hazard3 RV32IMAC (RP2350) at 150 MHz, on real hardware, same C kernels, clocks verified.

Feather M4 Express

Feather M4

Cortex-M4F (SAMD51)
Family
Arm Cortex-M
Board
Feather M4 Express
Clock
120 MHz
float / double
32-bit / 64-bit
Pi Pico 2W (RISC-V)

Pi Pico 2W (RISC-V)

Hazard3 RV32IMAC (RP2350)
Family
RISC-V
Board
Pi Pico 2W (RISC-V)
Clock
150 MHz
float / double
32-bit / 64-bit
Typical gap
1.19×
Feather M4 faster, geometric mean of 30 cells
Biggest win for Feather M4
6.24×
float divide
Biggest win for Pi Pico 2W (RISC-V)
1.72×
double modulo
Clock for clock
1.48×
Feather M4 needs fewer cycles per op (1.25× clock gap)

Operation by operation

Every cell

How many times faster Feather M4 is than Pi Pico 2W (RISC-V), operations per second. Blue: Feather M4 ahead. Orange: Pi Pico 2W (RISC-V) ahead.

AddSubtractMultiplyDivideModulo
uint8_t 1/1.36× 8.98 M/s · 12.2 M/s 1/1.48× 8.98 M/s · 13.3 M/s 1/1.48× 8.98 M/s · 13.3 M/s 1.32× 7.01 M/s · 5.3 M/s 1.20× 6.36 M/s · 5.3 M/s
uint16_t 1/1.26× 8.98 M/s · 11.3 M/s 1/1.26× 8.98 M/s · 11.3 M/s 1/1.17× 8.98 M/s · 10.5 M/s 1.24× 6.12 M/s · 4.94 M/s 1.12× 5.36 M/s · 4.79 M/s
uint32_t 1/1.26× 9.71 M/s · 12.2 M/s 1/1.16× 9.71 M/s · 11.3 M/s 1/1.16× 9.71 M/s · 11.3 M/s 1.32× 6.53 M/s · 4.94 M/s 1.25× 6.19 M/s · 4.94 M/s
uint64_t 1/1.32× 6.19 M/s · 8.2 M/s 1/1.32× 6.19 M/s · 8.2 M/s 1/1.32× 5.61 M/s · 7.39 M/s 1.08× 885 k/s · 822 k/s 1.11× 885 k/s · 794 k/s
float 3.69× 8.98 M/s · 2.43 M/s 3.69× 8.98 M/s · 2.43 M/s 3.02× 8.98 M/s · 2.98 M/s 6.24× 4.73 M/s · 758 k/s 1/1.32× 866 k/s · 1.14 M/s
double 1/1.01× 1.2 M/s · 1.21 M/s 1/1.04× 1.19 M/s · 1.24 M/s 2.51× 1.7 M/s · 678 k/s 3.20× 1.26 M/s · 394 k/s 1/1.72× 425 k/s · 731 k/s

Where they sit in the fleet

Keep comparing

Swap sides