Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
One of the bigger cost items of implementing floating point in HW is the barrel shifter. Posit math makes that worse, because now the mantissa needs to be shifted before it can be used.
There are "up-to" a fixed number of exponent bits. Regime is RLE encoded. If the RLE encoding takes enough bits, exponent bits are lost.
The format, as described, has the most data points (geometrically) between 0.5 and 2.0 and -0.5 and -2.0. Very large (|x|>>2) and very small points (|x|<<0.5) have more regime bits, at most equal exponent bits, and fewer fractional bits.
The format you describe is not posit. It is some other variation of tapered floating point.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
