Thing Done

← The fleet

Teensy 4.0 vs Nucleo G474RE

Cortex-M7 (i.MX RT1062) at 600 MHz against Cortex-M4F (STM32G474) at 170 MHz, on real hardware, same C kernels, clocks verified.

Teensy 4.0

Teensy 4.0

Cortex-M7 (i.MX RT1062)
Family
Arm Cortex-M
Board
Teensy 4.0
Clock
600 MHz
float / double
32-bit / 64-bit
Nucleo G474RE

Nucleo G474RE

Cortex-M4F (STM32G474)
Family
Arm Cortex-M
Board
Nucleo G474RE
Clock
170 MHz
float / double
32-bit / 64-bit
Typical gap
7.03×
Teensy 4.0 faster, geometric mean of 30 cells
Biggest win for Teensy 4.0
35.5×
double add
Closest cell
4.20×
uint32_t modulo
Clock for clock
1.99×
Teensy 4.0 needs fewer cycles per op (3.53× clock gap)

Operation by operation

Every cell

How many times faster Teensy 4.0 is than Nucleo G474RE, operations per second. Blue: Teensy 4.0 ahead. Orange: Nucleo G474RE ahead.

AddSubtractMultiplyDivideModulo
uint8_t 6.59× 77.4 M/s · 11.7 M/s 6.41× 75.3 M/s · 11.7 M/s 6.41× 75.3 M/s · 11.7 M/s 5.84× 52.3 M/s · 8.96 M/s 5.71× 46.3 M/s · 8.1 M/s
uint16_t 6.41× 75.3 M/s · 11.7 M/s 6.39× 75 M/s · 11.7 M/s 5.84× 68.6 M/s · 11.7 M/s 5.06× 41.5 M/s · 8.2 M/s 4.96× 37.5 M/s · 7.56 M/s
uint32_t 6.56× 82.8 M/s · 12.6 M/s 6.79× 85.7 M/s · 12.6 M/s 5.43× 68.6 M/s · 12.6 M/s 4.51× 39.3 M/s · 8.73 M/s 4.20× 34.8 M/s · 8.3 M/s
uint64_t 8.35× 63.2 M/s · 7.56 M/s 8.35× 63.2 M/s · 7.56 M/s 6.56× 43.7 M/s · 6.67 M/s 4.78× 5.54 M/s · 1.16 M/s 4.78× 5.54 M/s · 1.16 M/s
float 6.62× 77.7 M/s · 11.7 M/s 6.62× 77.7 M/s · 11.7 M/s 6.62× 77.7 M/s · 11.7 M/s 4.21× 27 M/s · 6.42 M/s 4.98× 6.67 M/s · 1.34 M/s
double 35.5× 56 M/s · 1.58 M/s 34.7× 54.7 M/s · 1.58 M/s 19.7× 43.7 M/s · 2.22 M/s 9.39× 15.7 M/s · 1.67 M/s 7.04× 4.84 M/s · 688 k/s

Where they sit in the fleet

Keep comparing

Swap sides