Forum

Ben a écrit :J'ai introduit le même programme (le calcul de factorielle avec le méthode des logarithmes). Et bien, le FX-880P a terminé quelques secondes avant le FX-850P! J'ai été assez surpris.

Salut,

A possible reason for the difference could be the lack of the CLEAR command in your test program.

I'm sure that the label "FX-8000P" is a fake. The calculator itself seems to be a SHARP or a SHARP clone and has nothing to do with CASIO. The key layout reminds me to some of the SHARPs in the EL-5xx line, but I cannot say which one exactly. I have also noticed some non existent CASIOs in the list ...

Salut Ledudu, please allow me some corrections for your site: The "FX-8000P" is not a CASIO, but a SHARP in the EL-5xx line or a SHARP clone. The PB-1000C isn't programmable in C. The HD61700 assembler was replaced with a CASL assembler. The FX-840P is not prgrammble in native assembly. The FX-770P,...

Salut Daniel, merci for your kind words. The HD61700 is a highly interesting CPU with a very nice and special instruction set. Unfortunately the PB-1000 assembler contains a subset of these instructios only. What I miss particularly, are the instructions for LCD access and the multi byte commands, t...

Assembly language routine for the PB-1000 series for large numbers up to 16 digits in BCD format. Most of the instructions are not supported by the built-in assembler of the PB-1000. @00: ADW $24,#1 LDL $8,$0,L8 ADBL $0,$8,L8 ANC $8,#1 JR NZ,@01 ADBL $8,$0,L8 ADBL $0,$8,L8 DIDL $7,L8 XRC $0,#1 JR NZ...

Slow, slower, TI-62: 00 1 01 STO+1 02 x<->t 03 2 04 STO/0 05 RCL 0 06 x=t 07 R/S 08 FRAC 09 C.t 10 x=t 11 RST 12 6 13 STO*0 14 1 15 STO+0 16 RCL 2 17 x<->t 18 RCL 0 19 x>t 20 STO 2 Usage example: CM 27 STO 0 RST R/S RCL 1 -> 111 RCL 2 -> 9232 Execution time: 380 seconds

Not that easy, but possible on the FX-180P using 27 bytes and 4 vars: 01 2 02 Kin/1 03 Kout1 04 - 05 1 06 Kin+3 07 = 08 1/x 09 Kout1 10 Kin2 11 FIX0 12 RND 13 NORM 14 Kin-2 15 . 16 5 17 Kin+2 18 Kout2 19 x>0 20 6 21 Kin*1 22 1 23 Kin+1 24 Kout1 25 x<=M 26 Min 27 RTN Usage example: KAC Min 27 Kin1 P1...

Rounding: 01 FIX0 02 RND 03 NORM Integer part: 01 - 02 . 03 5 04 = 05 FIX0 06 RND 07 NORM Fractional part: 01 Kin1 02 - 03 . 04 5 05 = 06 FIX0 07 RND 08 NORM 09 Kin-1 10 Kout1 Round(x)-x to loop, if the fractional part is not zero: 01 Kin1 02 FIX0 03 RND 04 NORM 05 Kin-1 06 Kout1 07 x^2 08 x>0 Roun...

The fx-180p Plus (equivalent to fx-3900Pv) is faster and offers 300 steps in 4 areas with a nice program editor.
The fx-3900Pv needs 1.5 sec for 100 iterations.

Salut,
the fx-180Pa and fx-180Pv are identical function-wise, but the Pv is around 4 times faster.

A demo for self modifying code on the PB-2000C:
The key sequence modifies line 50 and starts the program again.
The code is only executable in the editor oder BASIC mode.

What an optimization compared to the 321 byte code. Well done Piotr!

Another solution is a self modifying code using the keyboard buffer, if available.

Salut,
LRAM and SRAM are available on many 4-bit casio pockets. It's for transfering the complete RAM content
(programs, variables and system area) to another pocket using a special cabel. More information about
the wiring and usage: http://www.pisi.com.pl/piotr433/fx730_e.htm

Marge a écrit :One question out of topic: did you run your benchmark test on the HP-29c? I didn't see any result for this machine....

Merci beaucoup for your offer, but the 29c is already tested with 42 minutes. I know the list is very long and sometimes it's easy to overlook a calculator.

It has already a Trace mode that displays the current step when running if you think it's too quick. ;) The TI-62 is a bit curious. As you mentioned there is a trace mode to keep track of a running program, but even in trace off mode you can still see the programming steps running but faster of cou...