Nintendo 64 and other later Gen Cores

[straddle]

Apologies if this has already been answered in the past - I only recently discovered the MiSTer project, and am yet to assemble my first build. I notice that Cores have been developed for an amazing range of 16-bit platforms, and a basic PC.

How feasible is it that we might see more advanced cores developed like the Nintendo 64, Playstation or any other later consoles? Is the DE10 Nano theoretically capable of recreating this kind of hardware? What other consoles might it be able to recreate?

[kitrinx]

Apologies if this has already been answered in the past - I only recently discovered the MiSTer project, and am yet to assemble my first build. I notice that Cores have been developed for an amazing range of 16-bit platforms, and a basic PC.

How feasible is it that we might see more advanced cores developed like the Nintendo 64, Playstation or any other later consoles? Is the DE10 Nano theoretically capable of recreating this kind of hardware? What other consoles might it be able to recreate?

Unlikely. There are several hurtles to overcome, including the volume of FPGA space such systems would take, the rate they would run at, the memory requirements they would have, the floating point operations they require, and most importantly, the enormous amount of R&D that a developer (or team of developers) would have to do to make them work.

The best you will likely see on the current iteration of MiSTer hardware is GBA and *maybe* PS1, but even those are quite far from reality, despite whatever tweets and optimism you may have seen.

Even if you ignore the hardware questions, you also have to ask: What is the advantage of an FPGA implementation on a 3D system that uses a framebuffer already? These systems can leverage modern GPU's to great effect to process and enhance 3d graphics (which arguably need enhancement a lot more than 2d), and with equivalent latency to the original system, unlike older 2d platforms which had no framebuffer. With our current understanding of the systems, does an FPGA even have the capacity to be any more accurate than software, or even equally accurate? The benefits get a lot more fuzzy.

[Sorgelig]

In theory N64 is possible, but probably FPGA size of MiSTer won't be enough.

[Dacide]

In this video RTL Engineering says that a n64 core "could very tight fits"

https://www.youtube.com/watch?v=6QbRuhe7WMQ

[Newsdee]

I'd rather see an X68K, PC98, PC-FX, and other big Japanese retro systems on MiSTer before a N64 and PS1

Now if only we could get the attention of the Japanese HDL developers...

[srg320]

IMHO, the main problem is memory.

N64: one chip 9bit @ 500MHz (inside RSP 64+8bit @ 62MHz).
Saturn: many SDRAM chips >= 2Mbit each, max 28 MHz, but with separate buses (2 chips for CPUs, 5 for VDPs, 1 for sound, 1 for CD cpu).

So, can't do without another additional memory.

[high5]

In this video RTL Engineering says that a n64 core "could very tight fits"

https://www.youtube.com/watch?v=6QbRuhe7WMQ

And in the end of the video he says it will take him at least 3-5 years to finish
I would stick to the more short-termed options.

[vanfanel]

What is the advantage of an FPGA implementation on a 3D system that uses a framebuffer already? These systems can leverage modern GPU's to great effect to process and enhance 3d graphics (which arguably need enhancement a lot more than 2d), and with equivalent latency to the original system, unlike older 2d platforms which had no framebuffer. With our current understanding of the systems, does an FPGA even have the capacity to be any more accurate than software, or even equally accurate? The benefits get a lot more fuzzy.

This is very, very interesting. I know for sure N64 uses a framebuffer, but does PSX also use one? I guess every 3d-oriented machine does, since the final rasterized frames have to be stored somewhere before being displayed (correct me if I am wrong here).
So, what about PSX and Saturn? If I understand you well, a PSX emulator with a true double buffer scheme (max_swapchain=2 in RetroArch, running on KMS/DRM with a MESA GLES implementation ensures such an scenario), we should get about the same inpug lag we get on a real PSX connected to the same screen, right?
As for the Saturn, would it be the same?
Looking at https://segaretro.org/VDP1_(Saturn), it seems that the Saturn relies on a double buffer scheme to blend the VDP1 and VDP2 graphics, etc, but I do not know how that translates into the video lag domain.

[jemismyname]

I'd rather see an X68K, PC98, PC-FX, and other big Japanese retro systems on MiSTer before a N64 and PS1

Now if only we could get the attention of the Japanese HDL developers...

I keep hearing about a japanese computer called "FM Towns Marty" that looks quite interesting. i would like to see that.

also, i would love to see sega saturn before i would care as much about playstation or N64... mostly due to the fact that Saturn games are completely outrageously expensive on the retro games market

[aristogeiton]

So install an ODE or buy a Satiator. Problem solved!

[aristogeiton]

Or PsuedoSaturn and a burner. Even cheaper!

[jemismyname]

pseudosaturn, eh? that's very interesting! thanks for that info. I am very likely to purchase a saturn now. i think the japanese consoles are still quite affordable.

I used a lot of pseudoephedrine to play saturn games all night when i was in high school. this will really take me back!

I'd still love to see a mister core, tho...
MiSTer all the things!!!

(although i'm sure saturn is the trickiest 90s console)

[SupraDeus]

someone is actually working on an fpga implementation of the psx:
https://github.com/Laxer3a/MDEC

What would be needed to bring this to mister?
I'm willing to do the work myself, but I need some pointers as to what it takes to write a mister core.
I have some experience with verilog and a lot with computer programming.

[Tibob]

I m sorry to be shelfish, but old arcade systems will be awsome instead of old first 3D consoles.

Cps1, cps2, system16, taito f3 and so many arcade system, sorry i m dreaming... or not

[crocky]

I m sorry to be shelfish, but old arcade systems will be awsome instead of old first 3D consoles.

Cps1, cps2, system16, taito f3 and so many arcade system, sorry i m dreaming... or not

You arent shellfish. CPS1 and System 16 might be coming to MiSTer in 2020 not sure about CPS2 and F3 looks far fetched (maybe never).

[onaryc29]

someone is actually working on an fpga implementation of the psx:
https://github.com/Laxer3a/MDEC

What would be needed to bring this to mister?
I'm willing to do the work myself, but I need some pointers as to what it takes to write a mister core.
I have some experience with verilog and a lot with computer programming.

I think you should wait for this port to be more complete or perhaps help/contact the current coder(s).

[SupraDeus]

someone is actually working on an fpga implementation of the psx:
https://github.com/Laxer3a/MDEC

What would be needed to bring this to mister?
I'm willing to do the work myself, but I need some pointers as to what it takes to write a mister core.
I have some experience with verilog and a lot with computer programming.

I think you should wait for this port to be more complete or perhaps help/contact the current coder(s).

Good suggestion! I've reached out to the dev on facebook
Let's see..

[Chol]

Looking at https://github.com/Laxer3a/MDEC/blob/ma ... EV_LOG.txt the first line suspects that this is already targeted at MiSTer.

[Glaurung]

Looking at https://github.com/Laxer3a/MDEC/blob/ma ... EV_LOG.txt the first line suspects that this is already targeted at MiSTer.

I thought that was not possible due to the 3D chip.

[RabidWookie]

I m sorry to be shelfish, but old arcade systems will be awsome instead of old first 3D consoles.

Cps1, cps2, system16, taito f3 and so many arcade system, sorry i m dreaming... or not

You arent shellfish. CPS1 and System 16 might be coming to MiSTer in 2020 not sure about CPS2 and F3 looks far fetched (maybe never).

I don't see why CPS2 would be far behind CPS1, they're so similar in architecture.

[RabidWookie]

I m sorry to be shelfish, but old arcade systems will be awsome instead of old first 3D consoles.

Cps1, cps2, system16, taito f3 and so many arcade system, sorry i m dreaming... or not

Agreed, perfect versions of arcade boards with no frame buffer that are essentially unavailable otherwise are far more valuable than cores for framebuffered systems that can be explored already with $50 hardware and an ODE.

[Newsdee]

Honestly, this will boil down to whatever a dev has fun doing or is motivated to do

[djmartins]

Honestly, this will boil down to whatever a dev has fun doing or is motivated to do

Exactly this.
I am impressed with how many good cores keep popping up!
Also, anything retro that is very expensive in the market is a good motivation for someone to make a core if they can.

[FPGAzumSpass]

Keep in mind that this "psx" core you linked is ONLY for playing fmv videos.
It doesn't seem they really want to do a full psx.

On the other hand, i'm not sure why 3D would not be a goal.
Having a framebuffer does not mean it's unsynced to the screen as with an emulator, that normaly syncs to audio as it would sound like garbage otherwise.

Also a lot of "new and young" HDL Designers never have seen all these old consoles, for them a N64 can be nostalgic enough.

Winning the race to the top of the bob-omb-mountain or beating Diablo down in the cathedral on MY OWN core would feel so great, i doubt i can resist for ever...

[theimporter]

The best you will likely see on the current iteration of MiSTer hardware is GBA and *maybe* PS1, but even those are quite far from reality, despite whatever tweets and optimism you may have seen.

And 6 months later, a GBA core gets release. The future is NOW!!!