Why isn't there a flash cart for the Neo Geo?

[Joe Redifer]

I don't own a flash cart for any system, but with games in high demand and prices pretty high, I am surprised that a flash cart for the Neo Geo (AES, MVS or both) does not exist.  The games themselves are fairly tiny, don't use special chips like SuperFX, MMC or VR, don't use S-RAM and a great deal of flash memory wouldn't be needed. 

I am amazed that there is no interest in homebrew for this thing.

[PCEngineHell]

There was a game copier setup available for it long ago. It would only do the early games though.

[Tatsujin]

as for the NG, most people would like to go with the real deal, probably.

[SignOfZeta]

Yeah, I remember some sort of copier back in the day. It probably cost $500 and sucked.

I think the answer to your question is "because nobody wants one".

People who buy AES want to spend way to much fucking for something. They pay $400 for Samurai Showdown IV on AES when they could have the entire series on MVS for half that.

....and People who buy MVS are only paying $10-50 a game in most cases so they don't really care about it.

[Joe Redifer]

What about homebrew?  I would think this would be an awesome homebrew console.

[rolins]

There was a game copier setup available for it long ago. It would only do the early games though.

The copier was called MGD2 (Multi Game Doctor 2) it had all sorts of extensions to play backups on various consoles. It obsolete and useless since every game is pretty much been dumped. You'll find them mainly in hands of collectors.

Here the Neo Geo Interface for the MGD2. Massive cartridge that requires several special sram + dram cards which are unbelievably more expensive than the MGD2 unit.

[SignOfZeta]

Ah, good 'ol Bung.

[SuperDeadite]

Actually a lot of bootleg MVS carts on ebay are made with flash chips.  I believe some of the NeoForum members make Ghostlop MVS carts by using flash chips.  I think the real problem is the production costs.  Those NeoFlash PCE kits are like $150 right? And they are only 128megabit.  If you want say Metal Slug 3, you need 707megabit.  If you want a massive cart to hold multiple games, it would be really expensive I think.

[Tatsujin]

megabits aren't that expensive anymore nowadays

[SuperDeadite]

Mass produced would be cheap, yes.  But no one mass produces flash carts.  Low production custom hardware is not cheap.  Hell the homebrew Last Hope AES carts were $600, and you had to supply your own Cart Shell and Case.  All they gave you were the internal PCBs, manual, and insert.  Seeing as the DC version was about $30, that money must have gone somewhere right?

[Joe Redifer]

I can imagine a giant tower/cartridge holding up to 24 Neo Geo games.  If you barely touch it, the game would freeze and the graphics would get all FUBAR'd.

[ooPo]

There's a good description of the problems here:

http://www.tototek.com/phpBB2/viewtopic.php?p=5597#5597

Neo Geo is a very difficult system to make a playback device for, far more difficult than NES.

The issue stems from 2 obstacles:

1) The Neo Geo has five ROM buses which must be emulated which means a very tricky design. All other consoles except NES, which has two, have only one!

2) The Neo Geo has enormous ROMs! To emulate all but one Neo game (KOF2K3) you'd need 512M (32-bit bus made from 2x 16-bit flash chips) for character (graphics) memory, 64M (16-bit bus) program memory for the game code, 128M (16-bit bus) for audio samples and 4M (8-bit) of Z80 program memory. That adds up to 708M for the largest game -- that would require over $200 of NOR type flash memory and that's assuming that the font ROM memory is integrated into character memory!

Instead of using flash, one could use cheap old computer SIMMs (this is what my DIY Neo geo RAM cart is designed around) but this creates another issue--power! Computer SIMMs consume about 4 mA per megabit, which means a ~710 megabit RAM cart with the support circuits would consume nearly 3 AMPS (15W)! Compare this to any other copier which consumes around 5W at most! Powersupplys that hefty often cost $30-50 and would be bulky.

Neo Geo like NES also uses bankswitching, special character bus chips and encryption. These issues can easily be worked around but make the design more difficult because later games which have their own form of bankswitching must be hacked to use the "standard" bankswitching.

The Neo Geo console's graphics bus is different from MVS and requires a logic chip with many many pins, essentially any home "playback" (flash)cart will have to have a MVS->AES converter built into the design to emulate this additional component in AES carts. Think integrated Phantom-1/NEO Super Converter (normally valued at $250 alone!)

Encrypted games (everything after KOF98?) will need to be decrypted before being flashed since emulating encryption would be a dumb waste of resources but thanks to MAME sources this is pretty trivial (the only thing that's trivial about emulating Neo carts )

Another substantial issue is the two PCB needed for Neo Geo games, in small numbers I would imagine the boards being as costly as $50 for a set since they're huge!

Alltogether this project is so extreme, it's simply not cost effective for anyone to make.

[guyjin]

  wow.

[SignOfZeta]

Honestly, for several reasons, I don't really buy those explanations. Am I really supposed to believe that its so hard to come up with a 3W power supply (too bulky? For a Neo Geo owner? Yeah, right.) or a dual PCB? I really think that stuff could be overcome if people wanted it bad enough. Also, I think the earliest pirate MVS carts are only a year or two newer than the Neo Geo itself. Making a pirate cart isn't the whole picture, but its a part of it.

The most popular machines are always cracked wide open and pirated to hell and back. I'm sure it wasn't exactly a breeze developing the first DS flash cart, but since the DS is the most popular system in the world, its super total cracked.

The most do-able solution would be something, like the guy said, that ran off RAM and to just have what's in the flash loaded as needed. I'm not the person to do this my any means, but I think painting a picture of the Neo Geo as impossible to run back-ups on serves the Neo Geo mystique, and now that the system is finally dead it needs some mystique pumped back into it. Kind of like how fans of classic American muscle cars just can't come to grips with the fact that their shit was fine for its day, but its just slow as hell compared to the new stuff.

[OldRover]

The notion that it has huge ROM space isn't the issue. The issue is the five ROM buses. It could be done in software but it is more feasible to use five separate flash memory chips. A DS flash cart is a piece of cake, like most other machines. It's not mystique whatsoever...you simply cannot make a cost-effective flash cart for the Neo Geo.

[guyjin]

what exactly are multiple rom buses good for? as far as I knew, the NES had only one...

[SignOfZeta]

The notion that it has huge ROM space isn't the issue. The issue is the five ROM buses. It could be done in software but it is more feasible to use five separate flash memory chips.

So did the price of flash memory just shoot through the roof and I didn't notice?

[PCEngineHell]

The notion that it has huge ROM space isn't the issue. The issue is the five ROM buses. It could be done in software but it is more feasible to use five separate flash memory chips.

So did the price of flash memory just shoot through the roof and I didn't notice?

Cost of designing and fabbing the pcbs and cart casing, memory, obtaining whatever special chips are needed, paying people to design it as a whole,and paying to have it built in batches/bulk. It all adds up, and isn't feasible when you know the odds are not great you will be able to recoup and that the sale price you will have to gun for will be up there. The vast majority of Neo collectors are not into bootlegging and will not bite easy on such a product.

 Also,as with the bootleg stuff sometimes being buggy,chances are good the same would be the case on a flash cart. In the end, it could be done, its just simply not cost effective knowing what all is involved and the demograph you would be trying to sell the product to.

[ccovell]

what exactly are multiple rom buses good for? as far as I knew, the NES had only one...

The NES had 2 ROM buses, one for graphics, one for program code.  The pluses of having multiple ROM buses is that the overall speed of the system is greatly increased, as the CPU doesn't need to load graphics from ROM, [decompress it] and display it on-screen, resulting in a delay.  Nor would the CPU, graphics, and audio HW have to share the same address bus, resulting in a delay as each hardware unit gets its slice of clock time.

The NES was pretty quick at animating its entire screen in certain ways because it could rely on the MMC/mapper to do graphics bank swapping.  The Neo-Geo similarly has separate buses for its graphics ROMs, sample ROMs, so each unit can function mostly independently without lagging the CPU.

[guyjin]

The NES had 2 ROM buses, one for graphics, one for program code.  The pluses of having multiple ROM buses is that the overall speed of the system is greatly increased, as the CPU doesn't need to load graphics from ROM, [decompress it] and display it on-screen, resulting in a delay.  Nor would the CPU, graphics, and audio HW have to share the same address bus, resulting in a delay as each hardware unit gets its slice of clock time.

The NES was pretty quick at animating its entire screen in certain ways because it could rely on the MMC/mapper to do graphics bank swapping.  The Neo-Geo similarly has separate buses for its graphics ROMs, sample ROMs, so each unit can function mostly independently without lagging the CPU.

Nifty, thanks