This ridiculously beautiful calculator is not a production model.

It is not even a prototype of a production model.

It is a "labor of love" by Gábor Kovács, a Hungarian engineer who took it upon himself to, well, build some beautiful calculators. Including graphiCAL that, although the name might suggest otherwise, is not a graphical calculator: rather, it is a very capable conventional programmable calculator, designed and built in the best traditions of Hewlett-Packard's classic machines, with RPN (Reverse Polish) logic.

The case looks manufactured, but it isn't: It's 3D-printed. The keyboard layout is interchangeable. In fact, I already received an updated keyboard layout from Gábor, matching the latest revision of the calculator's firmware.

Speaking of which, the calculator is still very much in active development. Its firmware is easily upgradeable thanks to its USB-C connector that is also used to charge its internal Li-ion battery.

Internally, the calculator is powered by a PIC microcontroller, with its firmware written in its entirely by Gábor, in the microcontroller's native machine language. Talk about true old school!

There are still a few software glitches but by and large, the calculator works well and works reliably. I daresay it's already one of my favorites. I especially like its well-lit, bright OLED display, which is a pleasure to look at. The display is large enough to show the calculator's full 4-level stack, or as many as 8 program steps at any given time.

To test graphiCAL's capabilities, I used an existing program, originally written for the HP-25C, that calculates the logarithm of the Gamma-function with roughly 8 decimal digits of accuracy. The program works "out of the box", pretty much, except that the HP-25C has no label keys and does not requires an R/S instruction to end program execution.

This program uses the expanded version of Stirling's formula. It computes the logarithm of the Gamma function for any positive argument and for all negative arguments where the Gamma function is positive; the result is usually accurate to more than 8 digits of precision. The program uses only the stack and the Last X register, leaving the calculator's memory registers alone.

000     LBL A
001      1
002     x<>y
003      9
004     x<y
005    GTO 012
006     Rv
007      ÷
008    LAST x
009      1
010      +
011    GTO 003
012     Rv
013    ENTER^
014    ENTER^
015     Rv
016     Rv
017      2
018      ÷
019      π
020      ÷
021     √x
022      ÷
023     ln
024     Rv
025     Rv
026     ln
027      ×
028      +
029     x<>y
030      -
031     x<>y
032      x2
033     1/x
034      3
035      0
036      ÷
037      1
038      -
039     Rv
040     Rv
041     Rv
042      ÷
043      1
044      2
045      ÷
046      -
047    	R/S