Alternative option: With a few gigs of read-only memory, you could burn a game image onto it, and market a device dedicated to that single game.

You could just about eliminate any and all compatibility and piracy issues, you wouldn't require customers to have an existing console, and conceivably you could still profitably sell the game at $50.

It wouldn't be very environmentally friendly (you're forcing players to buy a new console for every game) but it could make economic sense.

The devices are certainly small enough to stack behind your TV; you just need a decent KVM switch designed to work with these devices.

waste of hardware as you mentioned. your idea would have been popular 7-10 years ago.

Depends on the case and the game.

I suppose you are right, if the game were facebook and you could buy a $50 dongle that has facebook and netflix people would probably only run that game.

I don't think the consoles would take a dent from something like this. If a game market evolved from this it would likely be niche and be similar to some of the smaller/toy-ish options you can buy already. I just can't see developers jumping ship from the big consoles (especially the way they now foster indie games) and making games for this lackluster (in terms of gaming) albeit cool device. If there was a way to make the device portable/cheap (some sort of small usb monitor)--that's a different story.

Graphics or simulation complexity -- I don't think it'd run Dwarf Fortress very well, for example.

In all fairness, by the colony's fourth year (given enough animals), multi-GHz machines with GBs of memory start groaning.

I need to convince pg that an open source Dwarf Fortress implementation is worth funding. Either that, or figure out how to build a cryptocurrency around it so that all of those Bitcoin miners can help simulate my fortress when its deflationary period starts.

Just pay them bitcoins to simulate your dwarves and your fortresses.

OK, but you could do iPad-style games. All you need is a cool input device. I'm thinking a webcam, and gloves with a specific bandwidth of color. Though the image processing might be too much.

Maybe for the input device just attach it to a big multitouch screen. Toss in a battery, wifi (maybe 3G?), ~32GB of flash memory. Then people could use one portable device and developers could write games and publish them to a central store for only ~1-5USD a piece.

I'm liking that Model A as a pocket-size emulation box for early games machines (specifically the SNES and ZX Spectrum).

"Substantially better than PS2, [possibly] about the same as XBox 1" is Eben Upton, the founders description of 3d performance. Theres an interesting interview on this here

http://foundationlibraries.blogspot.com/2011/06/cfl-podcast-...

He also talks about the original educational purpose behind it and how its expanded due to interest from the 3rd world and others.

That's on the very low end for consoles, all current generation consoles do much better than this. It sure beats an N64, but it won't run Halo.

Gosh times have sure changed.

At 700MHz, this thing is clocked at nearly 100x the ARM3 I started out my career on. And it has a vector floating point.

It would be awesome building a really stripped down kernel on this with a basic graphics kit.

I hop they get the numbers right, because they are going to get a lot of orders.

My point wasn't that this device would defeat them, but that it's possible to make computers that are this ridiculously cheap. Add a little more horsepower you've got a serious competitor... And it should still be pretty cheap.

It is on par with a WII.

Aside from the lack of a dedicated GPU, anyways (the Wii's CPU is apparently just over 700MHz).

It certainly kicks the pants off the Xbox 1, if that ARM core's architecture can keep up with a P3.

I think I'm right in saying that it won't; as awesome as they are, they're not known for high workload per clock cycle.