Hacker News new | past | comments | ask | show | jobs | submit login

I'm really interested in the architecture of this thing. I'm guessing ARM running a NES emulator (rather than ports of the games), since they've got such great experience doing that on the 3DS.

If it's hackable, this would be a fantastic gadget. As a project box alone, it looks really pretty. HDMI out, USB power, and controller inputs make it really appealing. Depending on headers and connectivity inside, hacking a USB port or SD slot for expansion would be a fun project.

There are already Wii controller accessory to USB adapters available. And at $9.99 for the controllers, those could be fun too (to be used separately on other platforms, like the RPi).

That's not to say I won't be getting one purely to play as intended, but I would be interested in picking one up to hack.




Informed speculation seems to suggest that this is a 3DS (or at least the bits of a 3DS required to emulate an NES).

All the games listed with it are available to buy for the 3DS, and the 3DS already supports the "save state" feature.

One notable missing game is Castlevania III which apparently gives emulators trouble.


Castlevania III in the US uses the MMC-5 mapper[0] which is relatively complicated as far as mappers go. Most modern emulators have no issue with it, but many cheap Nintendo-on-a-Chip clones have it[1]. Castlevania III in Japan used expansion audio through connectors that were not available in the US, and thus has enhanced sound compared to the US counterpart.

Emulation systems, such as the Retron 5, deal with Castlevania III with no issue.

[0] http://wiki.nesdev.com/w/index.php/MMC5

[1] To see a common NES clone struggle and die in the process of playing Castlevania III, see this video from Satoshi Matrix: https://youtu.be/OnPv1xHsbvg?t=13m56s


15 years ago at a dodgy trade show booth in Vegas, I bought a $25 Nintendo clone controller that had been retrofitted to contain the guts needed to play a few dozen classic NES games that were pre-loaded.

I'm pretty sure it wasn't emulation, but actual miniaturized Nintendo-clone hardware running the original (pirated) ROMs. I have a hunch this isn't emulating, but is instead playing "on the metal." The hardware (as the article mentions) is easy to clone, and as such should be insanely cheap to produce in volume.

The profit margins on these consoles is going to be great.


Eh, these things are hit and miss. I have one of those $20 clones that accepts the original cartridges. Performance is pretty bad; the sound is miserable.


> The profit margins on these consoles is going to be great.

That would be true if it was an emulator or "bare metal". The beauty of IP!


Get yourself a Pi3[0] + nes30 controller[1]. If you're really going for an authentic look you can get a NES-style case for the Pi too[2]. Pi3 even has built in-bluetooth so it should work wirelessly with the Nes30 controller out of the box.

[0]https://www.pi-supply.com/product/raspberry-pi-3-model-b-new...

[1]http://www.nes30.com/

[2]https://www.raspberrypi.org/blog/lego-nespi-case/


Good idea.

Im also thinking might be able to use authentic controller but cut of the connector and instead wire directly to the Pi's GPIO.


Or SPI. The controller is just a shift register.


> I'm guessing ARM running a NES emulator

Really surprised people expect it's anything more than one of these but with official licensing.

https://en.wikipedia.org/wiki/Nintendo_Entertainment_System_...


I'm leaning towards ARM just because Nintendo already has a huge supply chain and manufacturing complex geared towards this architecture.

They're extensively bought into the ARM ecosystem. They already have a NES emulator on ARM (3DS NES Virtual Console). ARM SoCs are so incredibly cheap (look at the low volume RasPI with its HDMI output and now WiFI). The Nintendo hardware engineering teams have extensive ARM experience. There will undoubtedly be a business requirement for platform security (wrt the bundled games). This is, from Nintendo's perspective, a solved problem on their portable console platform.

Over a long period and at high volumes, sure - a NES hardware clone might make for a cheaper BoM. But that's not where the cost is here. It's in the engineering, initial supply chain and manufacturing startup. Sticking with ARM keeps all of those costs down.

Build a lightweight OS (based on the 3DS/2DS OS), able to run the same the same binaries, with the same libraries and you've cut out a massive amount of software engineering work and expense.

Ship some stripped down 2DS/3DS hardware, and you've saved a lot of hardware engineering work and expense.

I'd actually be surprised if it wasn't extremely close to the 3DS/2DS in terms of architecture, minus the IO, battery and secondary ARM CPU.


Is millions of units really low volume?!

Raspberry Pi passed 6 million last year.


Because the hardware clones are not sufficient. You need to make a sort of menu system + deal with hardware connections.

Simple ARM chip + software emulator means you could just do everything one one chip without headaches. Plus it'll likely be cheaper.


Those are usually rolled into the hardware clones in various ways. e.g. https://en.wikipedia.org/wiki/C64_Direct-to-TV (single chip ASIC C64)

It could be software emulation, but that's slightly less reliable, and has a boot time which may not be negligible.


This video talks about the different approaches that existing mini-retrogame systems have used. Some use emulation, some actually have the NES hardware on a chip.

https://www.youtube.com/watch?v=esQNEvcHVuM


What if this is Nintendo's "Android" device that has been rumored




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: