An Invisible Threat - Floating Point Error

Since the invention of computers - and continuing until today - real number calculations have produced unreported and invisible errors, sometimes catastrophically.

Why? To perform calculations, computers use a floating point representation of real numbers as an approximation.

What’s the problem? Representing real decimal numbers with a fixed number of binary digits, as is common in our computers today, introduces computational errors that can accumulate uncontrollably, especially as calculations become more complex.

Know It

Knowing is the first step. True North Floating Point system keeps track of the error so that it’s always known.

Though floating point error cannot be eliminated, it can be managed - becoming known instead of invisible - using the True North bounded floating point system.

Control It

Take control of the accuracy of your calculations with True North Floating Point, which provides fail-safe representation by maintaining boundaries for the real number values represented.

Embrace the change that provides assurance in your calculation results.