Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
06/17/06 05:50
Read: times


 		 #118458 - On double precision floating point
Responding to: ???'s previous message
Kiran

I am sure you have heard of Raisonance C51 compiler. It has what you need. BCD double precision float numbers(10 digits mantissa) and also BCD long double (12 digits mantissa) precision. As for Erik's statement about suitability for number crunching, I think its for the user to decide on his situation; if it is acceptable to get his results in 20 seconds or 20 milliseconds !!! I was in a similar boat many moons ago and tried out Raisonance. Of course if the number of digits of significance matter, double precision is the way to go. To my knowledge (may be flawed) single precision yeilds 6 significant digits and double 15.

Jerson

List of 32 messages in thread
TopicAuthorDate
Calculating accurate Distance b/w            01/01/70 00:00      
   Nothing to do with Keil?            01/01/70 00:00      
      Keil-specific            01/01/70 00:00      
         Of course, but does that not mean that t            01/01/70 00:00      
            Eternities.            01/01/70 00:00      
               maybe differencial gps            01/01/70 00:00      
                  If it is differential gps ...            01/01/70 00:00      
               actually, it's just a ruse            01/01/70 00:00      
                  You Cynic!            01/01/70 00:00      
                     I guess you're right, Andy            01/01/70 00:00      
   haversine formula            01/01/70 00:00      
      I think the problem isn't the formula ..            01/01/70 00:00      
         basically stuffed            01/01/70 00:00      
   Over what range?            01/01/70 00:00      
   Cross-posting            01/01/70 00:00      
   So what was the big panic, then?            01/01/70 00:00      
      Clearly........            01/01/70 00:00      
         is that a problem?            01/01/70 00:00      
            Nooooo eric.            01/01/70 00:00      
               well, you often seem to be wound up enou            01/01/70 00:00      
                  it's the coffee, doncha know ...            01/01/70 00:00      
                     its the morons who want impossible thing            01/01/70 00:00      
            Good humor!!            01/01/70 00:00      
         and he does that because it works            01/01/70 00:00      
         Yet to be solved!!!            01/01/70 00:00      
            Write your own library!            01/01/70 00:00      
               Fixed-point math            01/01/70 00:00      
            Specification!            01/01/70 00:00      
               Maths            01/01/70 00:00      
                  shooting sparrows with RPGs and shooting            01/01/70 00:00      
                     Someone's being misled            01/01/70 00:00      
                     On double precision floating point            01/01/70 00:00      

Back to Subject List