Search found 654 matches

by deltarho[1859]
Nov 19, 2017 2:15
Forum: Community Discussion
Topic: [offtopic] pico seconds VS nano seconds ?
Replies: 1
Views: 29

Re: [offtopic] pico seconds VS nano seconds ?

one nanosecond = 10^-9 seconds
one picosecond = 10^-12 seconds
by deltarho[1859]
Nov 15, 2017 18:52
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

So, it would appear then that on those occasions when compiler optimised FreeBASIC BASIC is faster than my attempts to replace BASIC by asm is simply because on those occasions the compiler is smarter than I am. <HaHaHa> That makes sense because the number of opcodes that I am au fait with is a smal...
by deltarho[1859]
Nov 15, 2017 6:26
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

@jj2007 Yes, HERE It is a pity that the forum software does not show post numbers. From that post we have ( -gen gcc -s console ) elapsed Asm: 172ms 24250000 leap years found (inline) elapsed Asm: 175ms 24250000 leap years found (inline) elapsed Asm: 174ms 24250000 leap years found (inline) (repeate...
by deltarho[1859]
Nov 14, 2017 17:55
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

@jj2007

Your latest results bear little resemblance to earlier results where R2 was faster than Asm(function).
by deltarho[1859]
Nov 14, 2017 16:48
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

All this reminded me of PowerBASIC's Fastproc which is a function that does not use a stack frame. It has some restrictions but the asm function we are using here can converted to a Fastproc.

PowerBASIC went from 547ms to 331ms, getting closer to FreeBASIC with compiler optimisation.
by deltarho[1859]
Nov 14, 2017 15:41
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

I mentioned that I got no speed increase using Naked. I just tried it again and got 266ms whether I used Naked or not.

However, I was using -gen gcc -Wc -O3.

With gas I went from 439ms to 281ms using Naked which is amazing.

Perhaps -O3 is doing a Naked like tweak so using Naked here is redundant.
by deltarho[1859]
Nov 14, 2017 15:01
Forum: General
Topic: For the graphics gurus, perhaps <smile>
Replies: 9
Views: 197

Re: For the graphics gurus, perhaps <smile>

Good idea. The thing about the normal distribution is that it is a natural phenomenon. Our brains may interpret a cubic approach as unnatural but I am only guessing.
by deltarho[1859]
Nov 14, 2017 14:51
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

@jj2007

Following the Naked keyword should be cdecl, stacall or pascal. I mentioned on page 5 that I try it but there was no change in speed.
by deltarho[1859]
Nov 13, 2017 22:08
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

Sorry to interrupt the current conversation but, just out of interest, I ported the ASM function to PowerBASIC and was surprised at the difference. FreeBASIC (-gen gcc -Wc -O3): elapsed Asm: 266ms 24250000 leap years found FreeBASIC (gas): elapsed Asm: 439ms 24250000 leap years found PowerBASIC: ela...
by deltarho[1859]
Nov 13, 2017 5:57
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

@jj2007

When you refer to "my inline version" is that an assembler written exe or am I misunderstanding.
by deltarho[1859]
Nov 13, 2017 2:12
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

Since, algorithm speed, should be measured, without any other changes applied. (such as: compiler optimizations) Even Mr deltarho[...], agreed to that. MrSwiss and I recently had a difference of opinion: He favoured no compiler optimisation in tests and I did. It occurred to me that we may be at cr...
by deltarho[1859]
Nov 12, 2017 20:10
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

Considering the variation of machines used for timing in this thread it could be regarded as pointless for anyone of us to publish timing results in milliseconds, that is in absolute terms. Better, in my view would be to publish in relative terms where one particular timing was designated as the bas...
by deltarho[1859]
Nov 12, 2017 19:58
Forum: Tips and Tricks
Topic: LeapYear Function (Boolean eval.)
Replies: 126
Views: 1906

Re: LeapYear Function (Boolean eval.)

'And 3' makes Roland's contribution the fastest FB code.
by deltarho[1859]
Nov 12, 2017 18:33
Forum: Documentation
Topic: -pp compiler option
Replies: 36
Views: 579

Re: -pp compiler option

St_W wrote:There's practically no moderation on this forum so you can say nearly anything without having to fear any consequences.

What a sorry state of affairs.

Go to advanced search