When working with large floating-point numbers that are stored internally using 32 bits (type Float), a loss of precision may occur due to the limitations of the 32-bit format. For example, a value of 33,554,442.6789 cannot be stored internally as the exact number using 32 bits. A value of 33,554,440 will be stored instead. Therefore, if users were to write 33,554,442.6789 to a device, the value read would be 33,554,440. When greater precision is necessary, a 64-bit Double precision number should be used if supported by the driver and device.