[34S] Time diff. STO.00 vs. STO 00 ?

03142014, 05:01 PM
(This post was last modified: 03142014 05:04 PM by Sikuq.)
Post: #25




RE: [34S] Time diff. STO.00 vs. STO 00 ?
(03142014 07:18 AM)Paul Dale Wrote: The majority of the discussions about the numerical accuracy are stored in the old forum archives. There was surprisingly little discussion about this elsewhere. Pauli, The story behind your 34S is astounding; yes, it certainly has not been exposed to the extent I would have guessed. I woke up several times over night thinking of the 34S team running around in Australia and Germany seeking algorithm books, and also recalling my own attempts to make my own Newton routine faster. Sometimes I would simplegraph the function on the HP41C loop printer and count the roots. Does the user FIX setting in any way affect the actual step length used?  Sikuq PS: Pauli, is there any way you can test how many loops it takes to calculate the squareroot of say 1234567 on the 34S? Is the step length fixed or does it reduce itself closer to convergence (i.e. sign change)? And how did it calculate the initial guess? 

« Next Oldest  Next Newest »

User(s) browsing this thread: 1 Guest(s)