Brute simple speed check program

[Bill Triplett]

At 3.2GHz the PC emulator needs one second to do 1000000 loops.
On an old Samsung Galaxy Note 4, the Android emulator is about 15 times slower than the PC version.
How do the actual physical calculators score? G1? G2?
Please try this program with a million loops, and share your test results:

Code:

// Input the number of loops.
// Output is the number of seconds.
EXPORT LOOP(N)
  BEGIN 
  LOCAL S := 0, T, K;
  T := TICKS();
  FOR K FROM 1 TO N DO 
    S := S+1;
  END;
  T := TICKS() - T;
  RETURN T/1000;
END;

[rawi]

Quote:At 3.2GHz the PC emulator needs one second to do 1000000 loops.
On an old Samsung Galaxy Note 4, the Android emulator is about 15 times slower than the PC version.
How do the actual physical calculators score? G1? G2?

My G1 takes 54.92 sec, my G2 16.718. I think this is quite impressing.

[Bill Triplett]

Rawi,

Thanks.

That means the simulator running on a Samsung Galaxy Note 4 is almost exactly equal to the G2. As a comparison, I wheeled out a TI-89 Titanium with hardware 3. A similar program incrementing a variable needed ten seconds to do 1000 loops. Ouch.

Bill

[chromos]

There is old benchmark thread:
https://www.hpmuseum.org/forum/thread-11...04029.html

[John Keith]

(09-06-2022 01:23 PM)Bill Triplett Wrote:  As a comparison, I wheeled out a TI-89 Titanium with hardware 3. A similar program incrementing a variable needed ten seconds to do 1000 loops. Ouch.

On my HP 50g in approximate mode, 1000 loops take 2.46 seconds.

[Bill Triplett]

The HP-50g is charming because of the time it saves due to using fewer keystrokes, for most practical problem solving.

I just wish someone would create an update of the version 2.15 ROM without the IFTE bug. I keep a spare unit around running the 2.09 ROM, so IFTE works fine for me, but still...

Surely HP would not lose money, if they released the 2.15 ROM source code at this point in time. Some bright person(s) would be able to fix the bug, after some careful reading.