Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: discussion period begins, until Jan. 26: "natural interval extension": friendly amendment to M001.02

HMMM.  What about memory access bandwidth and speed, and
its impact on the overall speeds of
single versus double precision?

Baker

On 01/24/2016 12:05 PM, Richard Fateman wrote:

The (re) submission of these letters to the discussion
raised a question in my mind.

   I am puzzled by the WG 2.5 vote urging the IEEE 754 committee to
incorporate two facilities.

*  [double precision interval arithmetic] *_at the speed of _*simple
floating-point arithmetic.
* provide___*high-speed*_ arithmetic [for dynamic precision intervals]

What puzzles me is the statement of speed requirements in a standard of
this nature.
Was WG 2.5 thinking that the IEEE committee was derelict in its duties
in allowing
some or all of the 754 arithmetic operations to be provided in slow speed?








--

---------------------------------------------------------------
Ralph Baker Kearfott,   rbk@xxxxxxxxxxxxx   (337) 482-5346 (fax)
(337) 482-5270 (work)                     (337) 993-1827 (home)
URL: http://interval.louisiana.edu/kearfott.html
Department of Mathematics, University of Louisiana at Lafayette
(Room 217 Maxim D. Doucet Hall, 1403 Johnston Street)
Box 4-1010, Lafayette, LA 70504-1010, USA
---------------------------------------------------------------