The Linpack benchmark code measures floating point computation performance. It was developed as part of the Linpack matrix library. For more information, see the Linpack Benchmark FAQ.
This code has been modified only to compile and run on the Raspberry Pi and Mac OS X.
Compiling and Running
Compile the code with the command:
gcc -O4 -DDP -DROLL -o linpackc linpack.c -lm
This compiles the benchmark using double precision (-DDP) values and not unrolling the loops (-DROLL). Use -DSP for single precision values and -DUNROLL to unroll the loops.
Run the benchmark with the command: