Mux

Actually, it wouldn't. Depending on the features of the FPGA, it's relatively straight-forward to switch. All an eval-board is is an FPGA + memory + some communication glue and IO to make it useful. And they *would* be able to show off many cores as you could just upload them directly from PC. Most old console cores are out and used to be available on fpgaarcade. For more cores, check out opencores.org to see what people are writing. In short, an eval-board running a game makes a helluva lot of sense to show people you've got shit up and running before you're embarking on creating a custom PCB. -Mux

I'd pick Digikey, but hey.. I sometimes order parts from Mouser.. :-) -Mux

'As we enter the costly prototyping phase'... Erm, aside from BGA's and other hard to solder components, this only becomes 'costly' if you start doing overnight runs and invest in your own equipment. Even then, you're not looking at $2-3m. For less than a grand you can have a pretty decent board. Tooling and injection molding would be expensive but they managed to get that cost reduced. If prototyping is MORE expensive that the original $500k anticipated for tooling, they're padding out shit big-time. I've worked in environments with inhouse equipment and where PCB's were ordered overnight for the sum of about $1k/turn. The shouldn't have to. Plenty of places that'll do a turn in a few days but nowhere NEAR that. Even for 8+ layer boards. They NEED to show something that works. Like I said earlier, even if it's on a devboard. And puhleez don't try to pull the wool over peoples' eyes because it'll backfire massively. So here's the $64k question. If they DID have something (anything), they could have done a KS and gotten a lot more backers. What gives? -Mux

So is their issue that KS requires an actual working prototype? At least that gives a potential backer SOME peace of mind. I looked at the translucent renders in the hopes that the PCB could potentially be real but nah, there aren't even any connectors that line up. What I don't understand is why Steve Woita is involved with this and (potentially) get a bad name if this shit blows up. That leads me to believe that either a) they actually have something up and running but just suck at marketing or b) he's struck some sort of deal to let them take the fall and gets a cut or c) he's genuinely sold out and is blinded by a potential payout.. I hope it's option 'a' -Mux

Can they please stop creating sizzle reels and show something tangible, like something actually running on a devboard? Doesn't even have to be a custom one. Just get a DE2 devboard and show something. Anything. -Mux

I poked around their site a bit more last night and they had an audio block diagram up that was just *insane* and completely going against what they're after with an FPGA. Why have this uber sound system that can do ANYTHING (SID, FM, AY8910, etc)? The point of having an FPGA is that you just pick and choose a SINGLE block for the system you're trying to create, not a giant cobbled together one. With regards to prototyping, I prototype stuff on an Altera DE1 board. They're $250 or thereabouts with much more expensive ones out there. What bothers me about this is that they haven't even shown that. With a hardware engineer on-board, they could EASILY have prototyped stuff on a pre-made board and created a custom one. Not 'being at that stage' is complete bullshit. The hardware guy is working in parallel and has fuck-all to do with the case design aside from chiming in which connectors he can support. -Mux

I'll chime in on RVGS... What I don't understand is why, at this stage, they don't have a PCB or a prototype. If they've been beavering away you have to have SOMETHING. Block diagram, schematic, whatever.. They could easily put this to rest by just showing a PCB. One thing that worries me is that nothing has been mentioned about the tool-chain. For me that was the hardest part. I ended up creating a secondary board that connected to my PC to download data and relied on not-so-great free assemblers and what not. No debugger at all. What are developers supposed to use for RVGS? For each platform, they'd need a decent compiler / assembler / linker / debugger or they're just going not going to be able to attract devs. Period. I know there board is 100x more complex but a single screenshot or quick video of a manufactured / populated PCB and their credibility is restored, as long as it's not smoke and mirrors. my 2 cents. -Mux

I chose for an on-board speaker as most vga monitors don't have an audio in. It'd be trivial to add an audio-out though. Didn't mimic any particular cartridge format but modeled it around whatever made it easiest to route. As far as the tech is concerned, it's ancient, stemming from the early to mid 90s. Upside is that it's really easy to prototype as they still come in through-hole packages and you can pick them up on eBay for cheap. Downside is that they're significantly less powerful than current FPGA's and harder to find. -Mux

I've got a prototype that I did about 2 years ago with some ancient technology.. Nowhere near as complex but technically they're not the first to develop a new cartridge based console that's reconfigurable :-) Doesn't have as nice a case though.. z80 @ 6.25mhz (yes, that means programming in assembler) 8K SRAM FPGA with configuration prom on the cartridge so each game could theoretically have it's own graphics / sound hardware. The one and only core I did had 2 scrollable tile-based layers, 32 sprites @ 16x16 and 64 palettes of 4 colors. Onboard speaker + amp VGA out 2 standard Atari joystick pins. Total cost: ~$30 including PCB... -Mux