[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: BID vs. DPD performance



John Crawford wrote on Fr 23.06.2006 00:25: 

There have been many posts expressing opinions/intuitions - without any
actual performance measurements - comparing the expected, or hoped for,
performance of Decimal FP in DPD v. BID. So far, I believe that only
Marius Cornea has published actual performance measurements on real
library code on modern hardware, although Mike Cowlishaw has published
some measurements of pieces of a potential future library. 

I would welcome more performance measurements. 

The measurement, done by Marius Cornea should be
evaluated by other committee members.  We, SAP,
has been offering Intel the cooperation for the
evaluation as an independent software vendor.  
I would appreciate, if Marius Cornea or John 
Crawford can offer me, or IEEE754 committee the 
BID implementation source in order to evaluate 
the distributed data. 

I also would like to confirm, that the distributed 
data by Marius, concerning DPD performance is 
incorrect.    

I don't think that any of us need any more opinions/intuition.

I agree that, that the purpose of performance 
measurement is not relevant to the open decimal 
format decision.  

- Nobu

754 | revision | FAQ | references | list archive