12C Platinum vs. 48GX vs. 50g (Speed Test)

[JDW]

Would any of you know of complex and time consuming calculation that could be fairly easily run on the 12C Platinum, 50g and 48GX, for the purpose of doing a speed comparison between the 3 calculators?

I found an IRR example on page 63 of this PDF, but I'm not sure how to run it on my 50g and 48GX. Getting the IRR in that example takes just under 5 seconds on the 12C Platinum (purchased in 2015). That's barely long enough to make a good comparison with the 50g.

What I'd like to see is a calculation that would take at least 15 seconds on the 12C Platinum, which would make it easier to see the speed differences with the 48GX and 50g. Obviously, the 48GX would be slower than the Platinum but the 50g would be many times faster.

One more somewhat unrelated question... The back of the packaging on my 12C Platinum has a Copyright date of 2015, and the barcode date shows 03/14/19. It says Made in Philippines on the back of the calculator. I assume this is the newest variant of the Platinum. The keypad seems fine to me with good tactile feedback. I've read that there were good and bad keypads in the 12C series through the years. Does anyone know if my "2015 and newer" variant is considered one of the better keypads or how it compares with the 1981 original 12C?

Thank you!

[Erik-M]

For speed tests you can have a look here https://www.hpmuseum.org/cgi-sys/cgiwrap...i?read=700

I don't know anything about 12C keyboards

[Gamo]

I got three 12C

1. HP-12C+ (made in China used Two Coin batteries)
2. HP-12C (made in Singapore used Three Button batteries)
3. HP-12C Platinum (made in Philippines used Two Coin batteries)

12C+ got the worst keys press with keys delay and bounce often this is very annoying when input program but this one is the fastest computation speed.

12C Singapore version is the best feel to use with nice feeling with good key press. The speed is not fast but you will get originallity of this classic 80's calculator.

12C Platinum Philippines version is the best of all 12C Platinum I only own Philipine version and
this one keys is all good no missing keys and this one give you decent computation speed with very big programming capacity and option to use ALG mode if user prefer.

If you like to observe the speed by the use of the [IRR]
You can try this "Polynomial Equation Sover" using the HP-12C with "Cash Flow" function.

Solve: 3X^3 - 2X^2 + X - 11 = 0

Setup Steps: [FIX] 4
[CF0] -11
[CFj] 1
[CFj] -2
[CFj] 3
[IRR] .... keep compute .... then display -41.7483

Final Steps: 1 [%] 1 [+] [1/x] Answer: 1.7167

Remark: This will only find real root

Gamo 9/2020

[pier4r]

Would this help? https://www.hpmuseum.org/forum/thread-9750.html

for 100 iterations
~ 3.3s - hp 50g, userRPL
~ 5.5s - hp 48gx, userRPL


12c is absent or reported in the thread but not yet collected by me.

[grsbanks]

(09-30-2020 04:16 PM)pier4r Wrote:  12c is absent or reported in the thread but not yet collected by me.

The "N Queens" problem is possibly too complex for the 12C to handle. It certainly can't handle the "summation" test mentioned in that thread because it lacks trig functions.

[JDW]

(09-30-2020 12:12 PM)Gamo Wrote:  I got three 12C

1. HP-12C+ (made in China used Two Coil batteries)
2. HP-12C (made in Singapore used Three Button batteries)
3. HP-12C Platinum (made in Philippines used Two Coil batteries)

There must have been more than one version of the 12C Platinum made because the the woefully out of date Platinum Manual (dated 2005) talks only about a model with a single CR2032 battery. Mine has two batteries. That leads me to wonder if there have been differences in the keys among "Platinum" versions through the years.

[Allen]

(09-30-2020 10:43 AM)JDW Wrote:  Would any of you know of complex and time consuming calculation that could be fairly easily run on the 12C Platinum, 50g and 48GX, for the purpose of doing a speed comparison between the 3 calculators?

Why does your speed test actually need to be complex and time consuming?? If you are measuring speed, would it be practical and informative to use a loop to add a number to the X register (or if not the X register, some variable, since the stack/garbage cleanup functions on the 50g and 48gx would be an extra penalty that the 12c would not have to pay)?

I feel a simple loop would really tell you more about pure speed than a "complex" calculation. I submit a more complex calculation (e.g. one that used advanced internal functions - Like timing the sqrt(n) for n ranging from 1,000,000>n>1 ) would would need to be normalized according to each calculator's implementation of the "complex" functions. I could have misunderstood the question, also.

[JDW]

(09-30-2020 12:12 PM)Gamo Wrote:  Solve: 3X^3 - 2X^2 + X - 11 = 0

Setup Steps: [FIX] 4
[CF0] -11
[CFj] 1
[CFj] -2
[CFj] 3
[IRR] .... keep compute .... then display -41.7483

I certainly appreciate the exact entry steps! Very helpful! But...

The IRR solution performed on my 2015-edition 12C Platinum took less than 2 seconds. I am also unaware how to key that into my 48GX or 50g for the sake of doing a speed comparison among those calcs. My thinking is, if a calculation takes 20 seconds or so on my 12C Platinum, it might take a measurable second or so on my 50g rather than just being instantly displayed on my 50g. Being able to measure speed differences is what I seek here. And it would be best if that could be achieved with a useful calculation rather than just a special benchmarking calculation that really doesn't have practical use outside benchmarking. That way I can tell people, "If you want to do this kind of calculation, it can be done on all of these calculators and it will be much faster on this particular model too."

(09-30-2020 04:16 PM)pier4r Wrote:  Would this help? https://www.hpmuseum.org/forum/thread-9750.html

Unfortunately, I went through that entire 10 page thread. It does not help. In fact, several people in that thread point out the 12C lacks Trig functions making the content of that thread irrelevant to the 12C.

I will admit though that the first link in the 1st post of that thread is semi-informative about existing benchmark results. But it does not tell me how to key-in the Queens problem on a 12C.

What are the exact steps to keying in the QUEENS problem on a 12C Platinum?

[JDW]

(09-30-2020 11:37 PM)Allen Wrote:  Why does your speed test actually need to be complex and time consuming?? If you are measuring speed, would it be practical and informative to use a loop to add a number to the X register (or if not the X register, some variable, since the stack/garbage cleanup functions on the 50g and 48gx would be an extra penalty that the 12c would not have to pay)?

Can you provide some exact key-in steps for the loop you propose as a test? I would intend to key it into my 12C first, then 48GX and then 50g.

As to WHY I want this, I was sent the 12C Platinum by a viewer of my YouTube channel and I am doing an unboxing and review. I have a 28S, 50g, 48GX and Prime in my existing HP calculator collection, so presenting those in my video, showing (1) speed differences, and (2) differences in how to key in the same function on all the calcs, would add something useful to the video, I think. Toward that end, some USEFUL FINANCE CALCULATION that can be run on all those calcs would be idea, rather than just a specially contrived benchmarking calculation.

I look forward to your kind reply. Thank you!

[rprosperi]

(09-30-2020 09:24 PM)JDW Wrote:  There must have been more than one version of the 12C Platinum made because the the woefully out of date Platinum Manual (dated 2005) talks only about a model with a single CR2032 battery. Mine has two batteries. That leads me to wonder if there have been differences in the keys among "Platinum" versions through the years.

There are at least 5 'generations' of 12CP, likely more; among them:

1. Silver face, 1x2032, No Parentheses, Undo, BkSp functions (fastest model, as no Undo overhead)
2. Silver face, 1x2032, adding Undo, Bksp, Parentheses (slower)
3. Black face, 1x2032, 25th Ann Model
4. Black face, 1x2032
5. Black face, 2x2032

Key feel differences (and related things like rattle) for a given model tend to be related more to the country of manufacture, than simply to a model's particular generation. It appears the manufactured solution, materials, etc. are local decisions and not mandated by the buyer (HP).

For more details, see Jebem's excellent article with lots of references and a nice summary document on the many flavors of 12C here:

https://www.hpmuseum.org/forum/thread-5180.html

[Moderator: Added Parentheses info to the above list]

[JDW]

(10-01-2020 12:39 AM)rprosperi Wrote:  There are at least 5 'generations' of 12CP, likely more; among them:

1. Silver face, 1x2032, No Undo, BkSp functions (fastest model, as no Undo overhead)
2. Silver face, 1x2032, adding Undo, Bksp (slower)
3. Black face, 1x2032, 25th Ann Model
4. Black face, 1x2032
5. Black face, 2x2032

"At least" is correct terminology insofar as my "Silver Face" (i.e., "silver" at the top of the calc surrounding the LCD only, while the rest of the calculator is black) has 2 CR2032 batteries (packaging copyright dated 2015). Also, mine has the left-facing blue arrow underneath the minus (-) key, which allows me to press g and then - to backspace 1 character. The look of my 12CP almost perfectly matches the 12CP Emulator here, except for the fact the emulation lacks the logos to left/right of the LCD, and except for the fact there is a silly orange "OFF" above the ON key, which for the life of me I cannot figure out. Pressing ON multiple times turns the calc ON & OFF, so there's no meaning to press f and then ON like other HP calcs. As a result, I cannot see why HP printed "OFF" above the physical ON key. Can you?

[Allen]

(10-01-2020 12:33 AM)JDW Wrote:  Can you provide some exact key-in steps for the loop you propose as a test?

Eric's apt reply, first posted by Xerxes has both the keystrokes for a counting loop I suggested, as well as exact key sequences for HP 12c queens problem.

[JDW]

(10-01-2020 01:45 AM)Allen Wrote:  Eric's apt reply, first posted by Xerxes has both the keystrokes for a counting loop I suggested, as well as exact key sequences for HP 12c queens problem.

Thanks. I had examined that super-long document twice before, but after reading your reply, I went through it super slow and found the relevant section, which I shall post below so nobody else need read through that long document for 12C related info...

HP-12C
HP-12C Platinum
HP-12C ARM
----------

01 Clear Σ
02 8
03 STO 0
04 RCL 1
05 RCL 0
06 x<=y
07 GTO 58
08 1
09 STO+1
10 RCL 1
11 STO n
12 RCL 0
13 STO Nj
14 1
15 STO+2
16 RCL 1
17 STO 3
18 1
19 STO-3
20 RCL 3
21 x=0
22 GTO 04
23 STO n
24 RCL Nj
25 RCL 1
26 STO n
27 Rv
28 RCL Nj
29 -
30 x=0
31 GTO 43
32 0
33 x<->y
34 x<=y
35 CHS
36 RCL 3
37 +
38 RCL 1
39 -
40 x=0
41 GTO 43
42 GTO 18
43 RCL 1
44 STO n
45 RCL Nj
46 1
47 -
48 STO Nj
49 x=0
50 GTO 52
51 GTO 14
52 1
53 STO-1
54 RCL 1
55 x=0
56 GTO 58
57 GTO 43
58 RCL 2
59 GTO 00

Now I need to know when to Press ENTER. Maybe that's clear to some people, but not to me. I can complete steps 01 to 06 without pressing ENTER, but I stop at step 07.

Also what is step 27 “Rv”? No such button exists on the 12CP!

Next, what does STO with a + or - before the number mean? I press CHS before that number??

Next, isn’t “33 x<->y” incorrect? I see a white marked “x><y” key and then “x<=y” in blue, but there is no x<->y key!

All said, I don’t know how to key this in. Do you?

[rprosperi]

(10-01-2020 02:17 AM)JDW Wrote:  Now I need to know when to Press ENTER. Maybe that's clear to some people, but not to me. I can complete steps 01 to 06 without pressing ENTER, but I stop at step 07.

Also what is step 27 “Rv”? No such button exists on the 12CP!

Next, what does STO with a + or - before the number mean? I press CHS before that number??

Next, isn’t “33 x<->y” incorrect? I see a white marked “x><y” key and then “x<=y” in blue, but there is no x<->y key!

All said, I don’t know how to key this in. Do you?

When entering a program, you only need to press Enter when you see one in the program listing, the author has figured out the necessary logic and will include them where needed. This program happens to not need one due to having been carefully thought through to carefully manipulate the stack such that an Enter is not required.

"Rv" is the Roll down key - As there is no text down-arrow character, it has become conventional to show it this way. Similarly, programs for other machines use R^ to mean Roll Up (the 12C does not have this function).

STO+, STO-, STO* and STO/ combine the math operator and the STO into one instruction, so 5 STO+ 8 means add 5 to the value previously stored in register 8 and then store that sum back into register 8.

"33 x<->y" is like the above item - since text does not have a character with the < and > one above the other, this is shown as "<>" or "<->" in text listings, so the white one you describe as “x><y” is the function to use.

hth.

[JDW]

(10-01-2020 12:43 PM)rprosperi Wrote:  When entering a program, you only need to press Enter when you see one in the program listing, the author has figured out the necessary logic and will include them where needed.

Either the author made a mistake or I am a complete idiot because of this:

07 GTO 58
08 1

I can perform step 07 just fine but when I immediately press the "1" key in step 08, I get "Error 4" on the LCD of the 12CP! That is basically what I was trying to convey in my earlier post. I am using a real 12CP but you can try it yourself on the 12CP Emulator here.

You told me to not press ENTER unless those program steps specifically call for pressing ENTER. For that reason, I am not pressing enter after "GTO 58" and so when I perform step 8, the end result is that I am actually pressing "GTO 581" which results in Error 4.

Can you please explain how I should type in steps 07 and 08?

Thank you.

[Valentin Albillo]

(10-02-2020 04:51 AM)JDW Wrote:  

(10-01-2020 12:43 PM)rprosperi Wrote:  When entering a program, you only need to press Enter when you see one in the program listing, the author has figured out the necessary logic and will include them where needed.

Either the author made a mistake or I am a complete idiot because of this:

07 GTO 58
08 1

I can perform step 07 just fine but when I immediately press the "1" key in step 08, I get "Error 4" on the LCD of the 12CP! That is basically what I was trying to convey in my earlier post. I am using a real 12CP but you can try it yourself on the 12CP Emulator here.

You told me to not press ENTER unless those program steps specifically call for pressing ENTER. For that reason, I am not pressing enter after "GTO 58" and so when I perform step 8, the end result is that I am actually pressing "GTO 581" which results in Error 4.

Can you please explain how I should type in steps 07 and 08?

Thank you.

It's quite simple. You're gettin Error 4 because you're trying to go to step 581, which doesn't exist. You need to do the following:

First, ensure that you're keying in the listed steps In Program mode, so that they get recorded as a program, not in Run mode which would execute them immediately and that wouldn't work at all. You must be in Program mode.

Second, every destination of a GTO must be a 3-digit number, like 007, 234 or 053. So, to key in GTO 58 you must key it in as GTO 058, to make it a 3-digit address.

That's all it takes: Program mode and 3-digit destinations such that GTO 47, say, is entered ad GTO 047.

Hope that helps.
V.

[JDW]

(10-02-2020 05:41 AM)Valentin Albillo Wrote:  It's quite simple...

Now that it has been explained in detail, I see the simplicity of it! Thank you!

But to help others, here's a more foolproof code listing:

HP-12C
HP-12C Platinum
HP-12C ARM
----------

P/R (Enters Program Mode)
01 Clear Σ
02 8
03 STO 0
04 RCL 1
05 RCL 0
06 x<=y
07 GTO 058
08 1
09 STO + 1
10 RCL 1
11 STO n
12 RCL 0
13 STO Nj
14 1
15 STO + 2
16 RCL 1
17 STO 3
18 1
19 STO – 3
20 RCL 3
21 x=0
22 GTO 004
23 STO n
24 RCL Nj
25 RCL 1
26 STO n
27 Rv (R with down-arrow key)
28 RCL Nj
29 –
30 x=0
31 GTO 043
32 0
33 x<->y (the white-silked x><y key)
34 x<=y
35 CHS
36 RCL 3
37 +
38 RCL 1
39 –
40 x=0
41 GTO 043
42 GTO 018
43 RCL 1
44 STO n
45 RCL Nj
46 1
47 –
48 STO Nj
49 x=0
50 GTO 052
51 GTO 014
52 1
53 STO – 1
54 RCL 1
55 x=0
56 GTO 058
57 GTO 043
58 RCL 2
59 GTO 000
P/R (exits Program Mode)
R/S (Run Program and displays "running" for at least 21 minutes on 12C variants and displays 876.00 when finished.)

My HP-12CP completed the program in 23 min. 14s.

Of course, now that I have done this, I see the 50g executes it in about the same amount of time, while the older 48GX does it about as fast as the Prime, making this speed test not much of a real-world test of speed among these specific calculators.

What I was hoping for was a typical yet complex calculation that could be done on the 12CP, 50g, 48GX and Prime to show the expected differences in processor speed; namely, that the Prime is fastest, followed by the 50g, and then the 12CP and ending with the 48GX. But it seems nobody has any ideas on that, so I guess I will give up on that idea. What I can say in my video is, "if you want to keep your 12CP running for 23 minutes, here's how to do it."

I still would appreciate knowing why there is an orange "OFF" label silkscreened above the ON button on the 12CP even though there is no need to press f and then ON to turn it off like other HP calculators. You just press ON once to turn it on and then press ON again to turn it off. Why then the orange "OFF" mark?

[EdS2]

Is it possible to put a financial calculation into a loop, so you can run it 100 times to get a more measurable figure?

[rprosperi]

(10-02-2020 06:38 AM)JDW Wrote:  I still would appreciate knowing why there is an orange "OFF" label silkscreened above the ON button on the 12CP even though there is no need to press f and then ON to turn it off like other HP calculators. You just press ON once to turn it on and then press ON again to turn it off. Why then the orange "OFF" mark?

The label "OFF" is present to remind new users that the [ON] key is also used to power off, which may not be apparent to a new user, likely feedback from 12C users in the past.

Note that ALL of the printing on the bezel is in yellow, for the shifted functions on those keys. Since pressing [f] prior to pressing [ON] does not affect powering-off, HP decided to not add the cost of adding another process step for printing "OFF" in a 2nd color. I was not in those meetings in HP, but I was in those meetings at Sharp .

[brickviking]

Actually, for those crazy enough to have fonts that cover a large number of Unicode codepoints, Unicode U+2277 will show this: x≷y, though I don't know how many people here have that particular symbol. Arial Unicode MS has it, as does Cambria/Cambria Math and DejaVu Sans.

(Post 334)