Prizm fx-CG50 announced for spring 2017

[TeamFX Profile]

The fx-CG50 OS is very similar to the fx-CG20 OSes. I am a bit disappointed. Looks as if there are no secrets to unravel.
Apart from the RAM-address there seem to be no grave differences.

As I said before, this is an intermediate model.
But it's even worse with the new fx-CG500: http://www.casio.com/news/detail/casio- ... calculator

"Casio Unveils Next-Generation Graphing Calculator"
My goodness, this is just a white fx-CP400 ... nothing new.

They have clearly run out of ideas for the SuperH calculator line-up.
New ARM-based models should be around by early 2019 (or by 2020 at the latest).

[TeamFX Profile]

Did someone post a picture of the new Graph 35+ E with spreadsheet app (OS 2.10, model CY801B)?
Source: http://www.casio-education.fr/articles/ ... ntree-2017

[cakeisalie5 Profile Website]

New ARM-based models should be around by early 2019 (or by 2020 at the latest).

So you don't think they'll switch to Renesas' RX architecture ? :p

[TeamFX Profile]

Either Renesas RZ or embedded ARM processors from a different manufacturer. Renesas RX does not seem to be powerful enough.

[cakeisalie5 Profile Website]

Ah yes, sorry, Renesas RZ. I don't think they'll switch to ARM, it's american... :p

[critor Profile]

Actually, almost all 3rd party fx-CG10/20 game apps don't work on the new fx-CG50/Graph 90+E.
The problem is too many developers did hard-cord the VRAM address, instead of getting it on-calc through the dedicated pointer.

Out of 38 game apps, only 3 of them are playable on the new models :
https://tiplanet.org/forum/viewtopic.ph ... 45#p215166

One of the working ones is Cubefield.

Nemhardy from Planete Casio did fix a minor related bug occurring with the Cubefield app on the new fx-CG50 / Graph 90+E models.
(graphics were ok, but text outputs were missing, as the text functions used were referencing the old VRAM address)

Check the new models huge performances compared to the previous models :

fx-CG10/20 :


fx-CG50 / Graph 90+E - twice faster ! ;D



Some more accurate tests would be needed, but the new models performances seem to be on par with the TI-Nspire.

The shown Graph 90+E is not overclocked.
I've tried neither Pover nor Ptune2, and anyway Ptune2 developer said it wouldn't work out of the box.



Fixed Cubefield download : https://tiplanet.org/forum/archives_voir.php?id=903366 (single app compatible with all fx-CG models : fx-CG10, fx-CG20, fx-CG50, Graph 90+E)

Source and more informations : https://tiplanet.org/forum/viewtopic.ph ... 48#p215248

[TeamFX Profile]

Ah yes, sorry, Renesas RZ. I don't think they'll switch to ARM, it's american... :p

Renesas RZ is an ARM processor design.

Renesas now meets with new processors we have created and technically support: our RZ family, advanced chips based on the proven ARM® Cortex™-A9 CPU.

[critor Profile]

Gravity Duck, same author and same minor problem, has now been fixed too.

Enjoy !


Source & more informations : https://tiplanet.org/forum/viewtopic.ph ... 67#p215267

[sentaro21 Profile]

Gravity Duck, same author and same minor problem, has now been fixed too.

Enjoy !


Source & more informations : https://tiplanet.org/forum/viewtopic.ph ... 67#p215267

I am glad to be seen in add-in for CG10/20 working well.

Original Gravity Duck automatically overclocked,
If this is overclocked, the hardware of CG50 seems to have a bigger margin.

Pover and Ptune2 will work to x26 PLL or more.
The RAM test does not work in Ptune2, but will work other than it.

Would you be able to show me the initial value of CS0BCR and CS0WCR ?

[SimonLothar]

The OS checksum:
At 0xA0B5FFF8
0xA0020000..0xA0B1FFFF + 0xA0B5FEF0..0xA0B5FFF7
(As with fx-CG20 OS 2.02)