That's fast. The 2 minutes I got in 2009 is on the same machine I have now with Intel Core Duo 2 Ghz CPU.
Does the same benchmark work on the iPhone? On Fri, May 11, 2012 at 9:10 PM, Joey K Tuttle <[email protected]> wrote: > Update on my "bug report" - > > H=: %@>:@(+/~)@i. > 6!:2 '%. H 100x' > 444.125 > > Not too shabby! And no crash. In the previous instance, I let the machine > go into sleep, so it may have been some snag about having a long running > process when iOS was trying to calm the machine... > > On 2012/05/11 20:49 , Joey K Tuttle wrote: > > Yes, originally I used a 50 50 matrix. I've been looking for where I > > recorded timings for various older machines. So if you find them first, > > please send them along. > > > > My memory, increasingly faulty, is that 50 50 took about 7 seconds on > > the 360/50 (with a total of 256 Kbytes of RAM!), but that may have been > > the time on the Philadelphia 360/75 much faster with a huge 1 Mbyte > > RAM... This machine routinely supported 100+ users with 80 Kbyte > workspaces. > > > > Another bit of related trivia is that the Philadelphia 75 was a return > > from NASA and the 60s space program. Also that Dick Lathwell had the > > console panel from it for years, I think he donated it to a museum, > > perhaps at Waterloo. > > > > For comparison, I just did the 50 50 on my iPad and it takes 0.00956 > seconds > > > > Thanks for the reminders - your Hilbert one gives me the first chance to > > report a crash of J on the iPad. > > > > I tried it on my iMac > > > > H=: %@>:@(+/~)@i. > > 6!:2 '%. H 100x' > > 80.0178 > > > > and then on the iPad - but after a few minutes, the app crashed, hard to > > say what happened, not likely a memory issue since it takes considerably > > less space than 500 500 ?@$ 1000 > > > > On 2012/05/11 20:07 , Roger Hui wrote: > >> Somewhere I have recorded the times for the "JKT benchmark" on various > >> machines over the years. But didn't you used to use 50 50 matrices? > >> > http://www.jsoftware.com/pipermail/programming/2007-December/009015.html > >> > >> Your report brings to mind two anecdotes concerning APL and matrix > >> inverses. The first concerns Ken's doctoral thesis work, for which he > >> solved some matrices for Leontief's input-output analysis. I've > forgotten > >> the size of the matrices involved; it was something like 50 by 50 and > was > >> on the Mark IV. Ken must have been an ace programmer. The Mark IV had > a > >> 200 word memory (?). > >> > >> The second anecdote is this: > >> http://www.jsoftware.com/papers/APLQA.htm#Hilbert_matrix . > >> > >> > >> > >> On Fri, May 11, 2012 at 6:38 PM, Joey K Tuttle<[email protected]> wrote: > >> > >>> Wow! Is a response I've made several times about this new version of j. > >>> Marvelous that such a powerful application is so compact. > >>> > >>> Since I couldn't resist, one of the first expressions I entered on my > >>> iPhone 4 (not the faster 4s) was - > >>> > >>> 6!:2 '%. 500 500 ?@$ 1000' > >>> 18.2 > >>> > >>> Amazing! > >>> > >>> The graphs on the iPhone retina are beautiful. I need strong glasses to > >>> hold it close enough to appreciate them and then the small font is OK > too. > >>> My iPad is a 2 and now I wish for the retina display there too (not to > >>> mention the faster processor) - the inverse runs in 4.57 seconds on the > >>> iPad 2, also amazing. > >>> > >>> - joey iPa... > >>> > >>> On May 11, 2012, at 8:27, John Baker<[email protected]> wrote: > >>> > >>>> First impressions - wow! > >>>> > >>>> I ran through the install process on plain old 3g connections - no > >>> problems. > >>>> The app comes up . I was able to step through the initial demos > without > >>>> problems. The standard J graphics > >>>> look great on the iPhone 4. > >>>> > >>>> Is there anyway to change the size of the default session font? The > >>>> current size is bit tiny for my old fart eyes. > >>>> > >>>> -- > >>>> John D. Baker > >>>> [email protected] > >>>> > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
