Homebrew Would a Nintendo 64 Emulator be possible?

[nooby89]

Yeah

 

[sarkwalvein]

Great, then you only need to get drunk enough to perceive time 4 times slower and you get perfect n64 emulation.

 

[bennyman123abc]

Yeah

Great! We just need optimization if it's at the least possible! Time to study up on the N64's CPU and GPU!

 

[nooby89]

This is false you know?

 

[huma_dawii]

Of course is possible, but nobody wants to take the challenge.. I wish I could code.. I would try to do a lot of things that are "impossible" to some people..

 

[sarkwalvein]

It is not impossible, but it would need to be done with a lot of hacks, not pure emulation but HLE tuned for each game like in the very old times of PC N64 emulation (UltraHLE).
It would be such a pain in the ass that I would shot my dick before taking upon such an effort (and I think that applies to us all).

 

[Joom]

It is not impossible, but it would need to be done with a lot of hacks, not pure emulation but HLE tuned for each game like in the very old times of PC N64 emulation (UltraHLE).
It would be such a pain in the ass that I would shot my dick before taking upon such an effort (and I think that applies to us all).

This is why I hope Nintendo starts releasing N64 VCs for the 3DS, but that's highly unlikely.

 

[Ace-D]

ctr64 is also death as of now. im wondering if anyone will pick up that n64 emulator project. its been sitting on github for at least 8 months now

 

[smileyhead]

ctr64 is also death as of now. im wondering if anyone will pick up that n64 emulator project. its been sitting on github for at least 8 months now

Wew, necrobump.

 

[DarthDub]

Super Mario 64 and nothing else.

 

[SoslanVanWieren]

get a gpd xd

 

[Platinum Lucario]

While it would be possible to implement Nintendo 64 emulation on the Nintendo 3DS. Unfortunately it would be limited in the capability of what it can do, if it was made as an app.

The best solution would be to develop an NUS_FIRM partition from assembly language, or some application that installs an NUS_FIRM partition onto the Nintendo 3DS system. Only that would provide optimal play and reach the limits of what the ARM11 and ARM9 processing cores can do. Since what we're handling here, is converting all the MIPS instructions into ARM9 and ARM11 instructions, which has to be implemented in the same way as AGB_FIRM is written to convert ARM7 instructions into ARM9 instructions.

But the thing is, the NUS_FIRM will have to be bigger, because all the instructions have to be converted, where as the AGB_FIRM only has to convert certain instructions of ARM7 into ARM9 (since most of their instructions are identical to each other).

I'm positive that NUS_FIRM could also be implemented on the original 3DS, but we should aim for New 3DS first, seeing as it has a higher clock-rate and more RAM.

 

[CatmanFan]

While it would be possible to implement Nintendo 64 emulation on the Nintendo 3DS. Unfortunately it would be limited in the capability of what it can do, if it was made as an app.

The best solution would be to develop an NUS_FIRM partition from assembly language, or some application that installs an NUS_FIRM partition onto the Nintendo 3DS system. Only that would provide optimal play and reach the limits of what the ARM11 and ARM9 processing cores can do. Since what we're handling here, is converting all the MIPS instructions into ARM9 and ARM11 instructions, which has to be implemented in the same way as AGB_FIRM is written to convert ARM7 instructions into ARM9 instructions.

But the thing is, the NUS_FIRM will have to be bigger, because all the instructions have to be converted, where as the AGB_FIRM only has to convert certain instructions of ARM7 into ARM9 (since most of their instructions are identical to each other).

I'm positive that NUS_FIRM could also be implemented on the original 3DS, but we should aim for New 3DS first, seeing as it has a higher clock-rate and more RAM.

Now that is what I'm going for. It would be great to see something like it.

 

[daxtsu]

AGB_FIRM only has to convert certain instructions of ARM7 into ARM9 (since most of their instructions are identical to each other).

AGB_FIRM isn't emulation (entirely), the 3DS has an ARM7, so the games run on their native CPU, with the cartridges and SRAM/eeprom/save flash being emulated, and if my memory doesn't totally fail me, the original DS GPU has 2D modes which are backwards compatible with the GBA, so it likely just directly uses that for picture processing (and then the ARM11 puts the actual picture onto the screen).

The best solution would be to develop an NUS_FIRM partition from assembly language

That would be a colossal task in and of itself, since there would no longer be an operating system running underneath. You'd (you meaning a person in general) have to effectively bring up and initialise all of the hardware yourself and provide drivers to utilise everything, basically reinventing the wheel and writing a miniature operating system. It made sense to do this for AGB_FIRM since you had a physical processor that you had to bring up (the ARM7), and you wouldn't want to risk it trashing the running operating system (not to mention the entire memory map would probably have to be changed). Since the N64 was also primarily a 3D system first, you'd need the aid of the GSP (the Pica200/GPU) to help out with pushing polygons, since software rendering wouldn't be an option on the 3DS, which would make things even more complicated. I'm not so sure that the amount of work needed would be worth the tradeoff of getting rid of the overhead that Nintendo's OS has; the 3DS simply has pretty old and weak CPUs (ARM11 came out in 2001, and the ARM9 around 1998-2000) no matter how you look at it, lacking things newer ARM chips have like NEON, which can help speed up some things quite a bit.