New Console: RETRO Video Game System

[Andromeda Stardust]

Wow. Trying to stay up on all the comments in three seperate forums is fun . To add more to the speculation the RETRO VGS won't be PC/Linux based nor Android based. We are going the FPGA route.

You say you're doing an FPGA. Care to elaborate what architecture this system is emulating? Generally FPGA are used to emulate custom chips, and due to retro console simplicity, an entire console can be hardware emulated on a single FPGA core. So is the FPGA microcode embedded in the game ROM, or in firmware? Generally people develop for a rigid platform that doesn't change. Developing for the FPGA would basically be passing the burden of developing both the hardware and software to the designer. I would think the extreme flexibility allowed would make programming it an extreme challenge. Or a homebrew developer could bundle an FPGA core for a classic console with a homebrew game designed to run on it. So the Retro VGS could literally be anything.

 

Call me intrigued.

[roland p]

I think the FPGA route is great since the developer is free to include a mix of any cores from the past in order to get a game with a retro (or arcade) look/feel.

 

Also, I like the idea of having a blank core, fed by what's on the cartridge.

[Keatah]

No reason we couldn't do mame-like accuracy. Or have modified cores excelling at the functions necessary for a specific game.

[Parrothead]

Here is some additional explanation from my partner, Steve Woita, "To answer the question: "What language will games be in? " My answer is: It depends. If a developer wants to make a Neo Geo game, they would include an HDL file that configures the FPGA to operate like a Neo Geo. The developer would code their game to run against the Neo Geo platform. This HDL code along with the actual Neo Geo game will be on the cartridge. Once that cartridge is placed in the RETRO VGS, it will become a Neo Geo and play that game. So in this case, the language is: 68000 and Z80 code. If you wanted to do a new 2600 styled game, you'd include a 2600 HDL file that configures the FPGA to replicate the logic of the original 2600 hardware and then you'd include your new 2600 game on that cartridge too. These two files are then paired up on the cartridge and when plugged into the RETRO VGS, will turn the console into a 2600. So the language that would be used in this case is: 6507 (6502 with less address space). Does that help explain things a little "bit" more?"

 

Note, we aren't doing this so much for hardware emulation of older software, but more importantly, giving developers the ability to program their new games using what they are already familiar with (Atari, SNES, NES, GENESIS, etc.). In a sense we might also be able to replicate an Android system and open this up to Unity developers as well. So basically, low level, Unity, etc. could all be possible. Also, it's important to note that cartridges are the real brains behind our console.

 

We are working on board layout, bill of materials, etc. so should have more info on this in the next few weeks.

 

 

 

 

You say you're doing an FPGA. Care to elaborate what architecture this system is emulating? Generally FPGA are used to emulate custom chips, and due to retro console simplicity, an entire console can be hardware emulated on a single FPGA core. So is the FPGA microcode embedded in the game ROM, or in firmware? Generally people develop for a rigid platform that doesn't change. Developing for the FPGA would basically be passing the burden of developing both the hardware and software to the designer. I would think the extreme flexibility allowed would make programming it an extreme challenge. Or a homebrew developer could bundle an FPGA core for a classic console with a homebrew game designed to run on it. So the Retro VGS could literally be anything.

Call me intrigued.

Edited April 15, 2015 by Parrothead

[Parrothead]

Exactly!

 

 

I think the FPGA route is great since the developer is free to include a mix of any cores from the past in order to get a game with a retro (or arcade) look/feel.

Also, I like the idea of having a blank core, fed by what's on the cartridge.

[Dan Iacovelli]

so by that explation its not just one system its a multiple game system. If code for the 7800(which is my next step after I retweek my 2600 game) and have it ported rvgs it will still be a 7800 game with the 7800 hdl file or if code for the jaguar the game would still be jaguar game with a jaguar hdl file. but can only be played on RVGS.

sounds good.

[Parrothead]

Yes.

 

so by that explation its not just one system its a multiple game system. If code for the 7800(which is my next step after I retweek my 2600 game) and have it ported rvgs it will still be a 7800 game with the 7800 hdl file or if code for the jaguar the game would still be jaguar game with a jaguar hdl file. but can only be played on RVGS.

sounds good.

[Hyperboy]

Now I'm getting interested, thanks for the updated explanation!

[Dan Iacovelli]

me too as well this system will realy help developers for homebrews.

[+remowilliams]

me too as well this system will realy help developers for homebrews.

 

Who are writing for the classic metal? Only if the FPGA cores are 100% accurate...

[+wongojack]

 

Who are writing for the classic metal? Only if the FPGA cores are 100% accurate...

 

Right, the last 2600 FPGA core that I tried was for the MCC and it was waaaay behind Stella in terms of accuracy. Where will the cores come from?

[+remowilliams]

Where will the cores come from?

 

My question exactly, writing these cores (at least if you intend them to actually function like the hardware they are intended to simulate) is not trivial.

 

Writing a bunch of new cores doesn't seem very likely at all. Using many existing cores out there will leave you with often incompatible or functionally incomplete simulation of the original hardware, which then means authors who are supposedly writing for the original hardware in this case have a nightmare on their hands.

[Dan Iacovelli]

if its a good core, just saying that aprogrammer can test a game on rvga if they don't have the tools like a flash card on actiual system before doing the game for the actual system

emultion on pc is fine but there is nothing like playing on actual system and posibbly have its own nich for rvgs

[Keatah]

Yes who will write new cores from scratch? Cores that are optimized and cater to the RVGS environment and intended use scenarios.

 

I don't know if I'm prepared to deal with all these nuances in such a new machine. I'm still recommending original consoles or emulation on a PC.

 

I like the idea of FPGA, but over the years it seems to have lagged behind emulators. And yet FPGA has so much promise. If you need extra FP performance you can get it. Maybe you need more voices, you can have them. And on the other hand. The PC platform is stupid powerful and pervasive everywhere.

[Dan Iacovelli]

well mike said he'll have more info in few weeks so lets take deep breath and exhale and wait until he come out with more details

Edited April 15, 2015 by Dan Iacovelli

[+CyranoJ]

Who's going to write, say, a Genesis game, and then put it in this thing as opposed to, say, a REAL genesis where it can reach 1000x larger audiences?

 

Trip Hawkings would be proud of this model....

[jaybird3rd]

First of all, as I understand it, the RVGS isn't necessarily limited to imitating older platforms. In addition to those, new cores can be created for new "virtual systems" which would fit into the style of vintage 8-bit and 16-bit consoles, or even something newer, but would support more modern development tools, making game creation easier. There are game designers out there (including students in my university's game design classes) who are interested in creating "retro-looking" games, but who are not interested in dealing with the difficulties of developing for 30-year-old hardware. Assuming the RVGS is successful, it may open up a new platform for them.

Secondly, the argument "but you can run the games on the original systems!" only holds if you actually have the original systems. If RVGS cores are produced for the NES, SNES, Genesis, 7800, and SMS, for instance, you'd need every one of those systems to play the games that could all be played on the RVGS. There's also the fact that those original systems generally don't have HDMI outputs, which makes them less convenient to use with modern displays, especially as newer TVs and monitors gradually phase out analog inputs. Once you buy one or two quality upscalers for your vintage systems, you're already at or beyond the (projected) cost of an RVGS console.

I can also see this being a great platform for multi-carts (load up a cartridge with a core for an old system and a bunch of ROM dumps or disk images), and there might also be potential educational markets for it.

What I like about this approach is that using a generic FPGA would greatly reduce the R&D costs of the console itself, leaving more funding available for the development of new cores. With certain exceptions, the cores for vintage systems that have been created so far (at least the ones I'm aware of) have been done by individuals as isolated projects, but with more resources, I'm sure that more and better cores can be created.

[Dan Iacovelli]

I understand about the tvs. you can't use s-video on but you can use composite or hdmi or rf for that matter

[Keatah]

Composite and hdmi is good enuf. If you need rf then you can get a modulator.

[Dan Iacovelli]

theres alsoan s-video to hdmi convertor you can get. I'm thinking about getting one for my jaguar untill I'll have it connected via rf on my 46 inch, but going to move it to an older tv

so I can have my 2600 and jag on one tv(right now I have jaguar and ota antenna connected for dtv broadcast)

[Shawn]

Here is some additional explanation from my partner....

 

That made things seem a little more interesting but for $200 you could have a WiiU which can emulate all the classic systems and play nintendo's modern offerings. Could be neat but I really, really hope you don't go bankrupt trying to get this thing to market as the whole idea just seems too out of touch with reality. I hope I'm just "not getting it" and you do great with this console and it sells a million units for you. Best of luck, I think you will need it.

[Keatah]

I don't think the focus of this project is tight enough. What does "RETRO" mean in this console's context? What does it mean to newcomers? What does it mean to veteran gamers?

 

Make no mistake. I'd like to see a cartridge-only system succeed against the headwinds of DLC and online accounts.

Edited April 16, 2015 by Keatah

[+CyranoJ]

Secondly, the argument "but you can run the games on the original systems!" only holds if you actually have the original systems. If RVGS cores are produced for the NES, SNES, Genesis, 7800, and SMS, for instance, you'd need every one of those systems to play the games that could all be played on the RVGS.

 

And for the cost of ONE game you could buy said system. And have far better accuracy.

 

This is never going to find people to make cores accurate enough. Its not going to find devs who wouldn't rather work with the original target hardware.

 

Nobody would rather make a 2600 game for this as opposed to a 2600. Same for <insert name here>.

 

While it was a dream of new h/w with a fixed set of features this was interesting. Now it's LOL FPGA on a cart... it's not going anywhere.

 

I do like the transparent cases though, keep making those.

Edited April 16, 2015 by CyranoJ

[+wongojack]

I know we are all a little skeptical, but on the positive side, this could have a preservation element to it. IF cores could be developed that were accurate then that would be yet another way that these classic systems could be preserved for the future on a relatively easy to use platform.

[+remowilliams]

Indeed, I've been a huge proponent of FPGA for exactly that - (potentially) immensely accurate and fairly easily used preservation of classic systems. And luckily it's happening to a degree, the MiST being one of the most well supported incarnations yet.

 

But that doesn't change the fact that the proposed model for this new console seems unworkable in several key aspects.