Jump to content
IGNORED

Retroblox


omnispiro

Recommended Posts

My first thought upon learning about this product was "how much are those modules going to cost and where am I going to store them when not in use?"

 

I also thought it might be a pain in the ass to keep switching between modules. Although I'm definitely not a fan of the Retron 5, I'd rather have a single console with multiple cartridge ports.

 

I honestly with the people from Hyperkin and Retro BIt would get their acts together and bring in some real experts. I believe the multi-console idea has great potential, just as long as the hardware is well built and it's not running buggy, stolen emulators full of input lag.

Link to comment
Share on other sites

Someone asked me in another thread to weigh in on "hybrid emulation", so here it is.

 

I don't think this will work out very well for a couple reasons.

 

First, the cart timing is pretty rigid, and you cannot "speed up" accesses to the cartridge past what a real system accesses it. i.e. for an Atari 2600 cart this is around 1MHz. While some carts might work, others probably won't. There will be a mandatory "waiting period" after setting an address before the byte can be read. Let's say it's 350ns. At 1MHz this means you only have 650ns left to process that CPU cycle. Let's say your ARM core is running at 1GHz to be fair. You've got 1000 cycles to use to emulate each system cycle.

 

This sounds like a lot, but if you're running ANY kind of OS at all, it will instantly be a failure. You flat out cannot have an OS of any kind running. In these 1000 cycles, you have to handle reading controllers (I hope they aren't USB!), output video (I hope you aren't using some kind of video library) and emulate your hardware. It must render the game/video/audio like a real system does, cycle by cycle. This means that in the 1000 cycles you have to run the audio once, the video once, and read the cart bus once. Then read controllers too. There cannot be drivers used in the usual sense, since they tend to operate in bursts (i.e. read the USB controller takes say 100uS which means you lost 10 cycles on your cart bus). Because you can't read the cart faster, you lost 10 cycles and can't get them back by reading the cart faster to compensate.

 

Then there's the fun of caching. the CPU is caching code/data and when you get a cache miss, it's going to stall things while it flushes/loads. This is going to steal a bunch of cycles at hard to predict times, and possibly cause more cart bus read fails. Trying to bang the cart via GPIO is going to be slow relative to the CPU- when it access GPIO it probably has a pretty large number of wait states which will also eat into your cycle count bottom line.

 

For systems with sprites like the NES, you're going to have to spend a lot of cycles processing the sprites each output pixel. In the case of the NES, pixels are generated at a rate of around 5.3 million a second. You must handle all 8 sprites (rendering with priority) that many times. For the 8 sprites+ backgrounds, this means 9*5.3 million pixels must be calculated and checked per second. That's nearly 50 million pixels. All of a sudden 1000 cycles doesn't sound like much. The NES has the double problem of running the CPU *and* PPU at the same time, and handling their interactions too. In short, there's just no way this can ever work except in the most trivial cases.

 

So those are the things I can think of off the top of my head. I'm sure there's lots more.

 

I rate success about as probable as the RGVS/Coleco Chameleon, except they did more work on the software end at least. I can guarantee they are not going to be writing their own emulators. It's just way too involved, especially for systems like SNES. Even NES emulation of any decent quality is a pretty big job that would take 3-6 months of 8 hour a day work to have it be anywhere close to accurate enough to run most titles/mappers. This is why I knew the Retron 5 was using purloined emulators. There was just no way to write that many good emulators in that short amount of time.

Wow Kevtris. Thanks for the input. So basically they are ultimately going to hit a brick wall and come to the conclusion that this "hybrid emulation" nonsense will not work, and default back to the Retron5 / Retrofreak way of dumping the game ROM then loading in an emulator. Although the CDROM port intrigues me (and it may be worth it to get it just for that), the only system they are currently advertising not supported by the Retron5 / Freak is Atari.

 

No NES homebrew, especially Memblers GT-ROM, RetroUSB UOROM, INL boards, or any other NES mappers with flash based saving, no flashcarts for any system, no Super Game Boy, SNES hacks using expansion chips may not run because the emu doesn't know how to run it, very doubtful that Melody based Atari homebrews will run, since most all of the embedded emulators for any system (Pi, Android, Wii, Xbox, PSP, etc) besides x86 Win/Mac/Linux PCs use an outdated version of Stella that doesn't support DPC+ AFAIK.

 

And that's assuming the Kickstarter is a success and they don't blow all their funding during development and actually deliver something. While I think add-on modules for Atari or other systems would be a good idea for existing product like Retron5, if the Retroblox isn't any better than the way the existing Retron5 / Rfreak does it, then it really isn't worth the money except for the CDROM support, which won't run anything without copyrighted BIOS dumps.

 

Kevtris, maybe you should rebrand the NT Mini as the "Zimba 2000" and sell adapter modules for it with laser cut acrylic shells (way cheaper startup cost than injection molding). I also bought a couple kits from Retrobuilt games, and am impartial to the wood aesthetic. Laser cut woodkits for the Zimba console would be the bomb IMO. You could your own Kickstarter to get the funds to have PCBs fabbed and buy laser cutting equipment and wood/acrylic supplies for the enclosures for your Zimba 3000 as well as cartridge modules. Maybe by the time it launches, you'll have the SNES/TG-16/Genesis cores ready.***

 

***Oops, wrong thread! :grin:

  • Like 1
Link to comment
Share on other sites

But their emulators "function at a much lower level than the ones you can simply download for your PC, Phone or RetroPie / RetroArch / LibRetro box." Nevermind the fact that these apparently master coders weren't able to get N64 and Sega Saturn working. Must just be a coincidence that these are the exact same two systems without 100% working emulators. :D

  • Like 3
Link to comment
Share on other sites

I thought they said they were doing high level emulation? Which in the traditional meaning is lame.

 

 

 

 

Hybrid emulation is a completely new way of tackling emulation, providing the cost-effective and future-friendly advantages of high level emulation as well as the accuracy and compatibility advantages of using clone hardware or FPGAs (Field-Programmable Gate Arrays) to model the original consoles.

Hybrid Emulation is unique in that it requires software emulators to function at a much lower level than the ones you can simply download for your PC, Phone or RetroPie / RetroArch / LibRetro box. RetroBlox’s proprietary RBXOS and Richter UI environment are built from the ground up in Linux, fully optimized for our system so that we can access the architecture from the bus-level and I/O from cartridge interfaces and controllers directly via high level emulation on the heart of the system, the Rockchip RK3288. This means we don’t have to rely on software emulation of memory mappers and/or co-processors that may not be fully optimized, which in turn frees up valuable resources to ensure other aspects of the machine run smoothly.

 

Sounds like they're trying to claim they can have their cake and eat it too.

Link to comment
Share on other sites

But their emulators "function at a much lower level than the ones you can simply download for your PC, Phone or RetroPie / RetroArch / LibRetro box." Nevermind the fact that these apparently master coders weren't able to get N64 and Sega Saturn working. Must just be a coincidence that these are the exact same two systems without 100% working emulators. :D

 

Even if their statements were true, low level emulation on a cheap, low powered processor, the same one that powers many phones and Chromebooks, isn't going to compete with a full-on PC.

 

Sounds like they're trying to claim they can have their cake and eat it too.

 

 

Affordable

Powerful

Compatible

Easy

 

If they're very good, maybe they can have two items on this list. One seems more likely. Having all four is self-contradictory in 2017.

  • Like 3
Link to comment
Share on other sites

Thing about this alleged system is...

 

I dunno. Why does this pie in the sky crap get my ire up so badly?

 

I would love some kind of magical, all in one solution that allows me to play my stupidly large collection of games with the real controllers (or my preferred controller, at least) on my big screen tv, in as accurate of a way as possible, with the added benefits of applying patches and hacks. "But, a pc and roms!" some of you may cry. Yep, that's a possibility... but when I sit there and look at a giant list of games, I'm often paralyzed by choice, if that makes sense. Too many options means I might only play something for a few minutes and capriciously move on to something else, a slave to whim. Using my physical carts/ discs means I have made a commitment, however vague, to play something for more than a few minutes. Using real games means ease, in a lot of ways, at least in my house hold. I have a rom solution thingie for my NDS, which lets me play 99% of any Game Boy, Game Boy Color or even DS games... and I don't use it much beyond playing games I would never be able to find or afford. When it comes time for me to really play a Game Boy game, it's usually on my actual GBA or on my Retron if my ailing eyes need the help. A flashcart full o' roms is neat, but not always the best solution. YMMV.

 

This is why projects like this one frustrate me so much; I hate having the hope swell for a nanosecond and then it's all dashed away by reality. Not that someday, a project like this won't be possible (I truly believe it will, and frankly, it will probably be necessary as more and more Duos and SEGA CDs and other temperamental machines die their final death)... but when someone comes along, with a potentially useful solution and instead, it's the same smoke-blowing hokum that I guess we as a community have to get used to, I find it both disappointing as a whole, and irritating in singular. The small advantage this thing may afford over a Retron or Retro Freak, the CDs, is not explained... we don't know about proprietary BIOS stuff (perhaps they don't, either... a scary thought), and they promise a new type of emulation which certainly just sounds like nothing all that new. It's still rock 'n' roll to me.

 

IT JUST WORKS. WAIT FOR THE KICKSTARTER. DON'T ASK QUESTIONS. DON'T YOU LIKE RETRO GAMES?

 

Let's not even get started on the Kickstarter thing... and I say that as someone who uses Kickstarter A LOT.

 

One day, Kevtris will come along on his gleaming white retro gaming horse and throw us a life preserver. Until then, I remain skeptical, and probably (okay, definitely) cynical... but there's still that ten year old in me that wishes he could play the good NES games and the good Atari games and the good SEGA games all on one box.

  • Like 4
Link to comment
Share on other sites

I don't get too excited about these projects. There's only a few that caught my attention, and even less that've panned out. I'm quite happy and content to putz along with existing software emulation.

 

As far as having a list of roms on a menu or flashcart. That shouldn't be a problem. Just pick one and play!

 

Having all systems in one box is doable, or very close to being such. It'll be interesting to mount the z3k boards (along with x86 and R-Pi) all into one case. Just a passing thought. But that way you could probably get complete coverage of anything classic. You could then have a customizable balance of fidelity and performance, convenience and versatility, all on modern hardware. All in one box. No need for chasing after a kickstarter and all the waiting and worrying.

 

Start now, build your X86, and tuck the R-Pi in thatspace between the mainboard and the bracket on the side. Don't worry, it'll fit.

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...