Benchmark from Tsoutsouras et al. MICRO paper.¹

The benchmark reads samples from a file, loads them on a distributional variable and then performs multiplication of the variable with itself. The product of the multiplication is a distributional variable.

double-multiply-autocorrelation -a <samples file> -m <mode>
	-a <samples file>: set to `input-A.txt` by default
	-m <mode>: 1 for explicit computation, 0 for implicit uncertainty tracking (0 is the default)

The samples are stored in a text file. The first line of the file is the number of samples that follow.

When running the implicit uncertainty tracking, the result may not be what you expect. Do not forget to check the underlying distribution.

With the implicit uncertainty tracking, the variable A gets a floating point value and an associated distribution. What gets printed is the value squared. The distribution squared is displayed as an histogram.