it's fun to play on the F. P. G. A.

[Keatah]

Thing with Polysmegma and Playmarjorie is that you have, what, 10? 20? ..systems being done by one console and essentially one small group of developers/integrators. That is a lot of work. And projects that ambitious simply don't work with a small crew.

 

And then you have things like Analogue NES/SNES clones. I believe the hardware to be capable of more vintage systems, but there's one man drawing the map. So things take time on that front. And on the other front, you have profit/commercialization. That means you can't give the store away either. These two factors alone will limit how far such consoles will go..Commercial products will only be produced if the money keeps flowing and the developers remain interested. Since most commercial stuff is closed source, a product stagnates if original interest fades.

 

Now enter open-source projects with many developers and plenty of teamwork. MAME is an example, many devs have come and gone, and the project rolls on. Anyone with skill can contribute. And there is always vibrant activity taking place. These projects also tend to develop great (vast) functionality with time. These projects benefit from many viewpoints. These projects tend to be immune to one dev calling it quits.

 

And we need qualities like that (and others) to begin to even think of preservation and longevity. I see many open-source projects outliving commercial & closed-source "efforts". And OS is very easily distributed for minimal/no cost. Remember proper preservation means providing access to all who seek. Software Emulation does this in spades. Hopefully with a low-cost platform FPGA can take part too.

[Keatah]

Ahh one other thing. Everybody seems to keep looking to these kickstarters and ramblings-on of projects that are nowhere near materializing.

 

To that I say pfagggh!! We have Software Emulation (established) and FPGA (getting underway) to pick from, and several commodity hardware choices to run them on. So. Just go ahead and build something up! Or get someone to help you if it's all too technical.

[Newsdee]

That's why these open source projects will have longevity over commercial solutions. It's not theoretical; MiST/MiSTer themselves are "descendants" of other projects like Minimig.

[phoenixdownita]

It's a nice write-up but MiSTer needs to have a believable option in the low-lag HDMI space.

I can't get over SmokeMonster mentioning repeatedly the low-lag option but via CRT ... which is all true so kudos to him for keeping the record straight, just that it should not be the case ... and yes it will come so we'll see.

I don't think it'll ever get commercial in its present form, those Terasic boards cost less that the FPGA they use so at that price point no one could make a commercially viable option. Also I am not too sure how many of those open-source cores are GPLv3 which commercial entities can't really handle too well for a host of reasons. So in the end I do see commercial entities selling cases/kits and specific add-on boards but the big elephant in the room (the DE-10 nano and its Intel Cyclone® V SE 5CSEBA6) it's not gonna be addressed.

For an hobbyist project is still plenty to enjoy, but for people awaiting a plug-n-play kind of a deal it may take a little longer, it wasn't so for the MiST as that board was an independent design (no subsidies in the form of below-cost prototyping boards).

 

If the price of the MiSTer main FPGA (Intel Cyclone® V SE 5CSEBA6) would drop sensibly, say by 50%, then the tides would turn for sure!

 

At least the core being open source may allow their preservation in source form, depending on the open-source license applied though it still may be complicated to get to a commercially viable option. It's a great thing that it is happening for sure!

[Keatah]

So what you're saying, essentially, is that FPGA popularity would really take off if we had sustainable low-cost hardware with which to make a standardized platform?

[phoenixdownita]

Not really, only that the 130US$ intro cost to MiSTer is only possible because of an heavily subsidized developer board which hides a 230+ US$ true cost of the FPGA alone (at least if an I7 spec component is really needed rather than the cheaper 150US$ C8 version).

In order to get from it (current MiSTer onto DE-10 nano) to a turn-key solution the main component of said board (Cyclone V 5CSEBA6, I have not even checked the cost of whatever else is essential) has to become way more affordable or the current cores need to be downsized to what the marker could bear for a commercial version of MiSTer. Or people are fine paying 300US$+ for a commercial version of the MiSTer (maybe they are, I do not know).

Keep in mind that just to offer HDMI on a commercial setting a company has to pay 10K annually to the HDMI consortium, which is nothing if you have the market but quite expensive for more niche products.

[youxia]

I'm not sure what is the point of these calcualtions. You can buy the base board for ~130$ right now, adding bells and whistles still keeps it under 200$. There does not seem to be any shortage, so what does it matter if these are subsidised or not? A commercial version for ~300$ (sum which could of course be lower too, given wider reach) - would still be leagues below the price of the nearest comparable product which is Nt Mini. And Nt Mini, as we all know, is either unavailable or has scalping prices starting above 1000$, which is nothing short of insane, if you ask me.

 

And MiSTer is already a low-latency option. The frame or two of difference vs 0 on Analogue device is only important to a very small group of people, it has no bearing on wider audiences, most of which already put up with lag in range of 6-10 on all the other retro machines. Even so, being open source, the chase to bring it even lower by coding better scalers is on.

 

Hopefully the positive buzz around it will continue and translate to 1:1 cores, more expandability and ultimately a boxed plug and play product.

Edited February 24, 2019 by youxia

[phoenixdownita]

I'm not sure what is the point of these calcualtions.

....

ultimately a boxed plug and play product.

The latter can really only happen if they can use a cheaper variant or the price of the original part drops considerably so they can build their own board.

I can't say for sure but I doubt Intel is really interested in subsidizing at a rate of 100US$+ apiece a hobby FPGA retrocore multi system, they sure as hell would like to be in on it but not by losing money, they can discontinue the product at any time, change the license etc...etc... and the problem originates from the fact that instead of designing a board with the needed parts MiSTer piggybacks on something else. It's a great and smart way to start (dev boards were intended exactly for cheap prototyping/experimentation) but sustainable I would not call it.

 

MiSTer as a hobby/enthusiast project is simply awesome, as a turn-key solution not so much.

It's great that the cores are open sources so they can continue to be perfected by anyone that has the skills, time and dedication to do it, no question about it.

[Newsdee]

The open source FPGA world is not monolithic.

 

MiST (not er) is a good alternative for anybody who feels strongly about having an independent project - i.e. which does not depend on a company product. It only has analog video output and a smaller FPGA, sure, but it is still being updated and has quite a lot of cores as well.

 

For a cheaper (but less powerful) alternative you can google for the ZxUno, it's a tiny FPGA board designed to fit an RPi case. It needs more tinkering than MiST and MisTer to update though, and it does not have any 16 bit cores (last I checked).

[phoenixdownita]

I though MiST was all but dead in terms of HW .... anyhow thanks for posting news about both projects.

[Newsdee]

I though MiST was all but dead in terms of HW .... anyhow thanks for posting news about both projects.

Not really, the MiST hardware is still sold new and there are developers working on cores. Some enhancements are ported over from MiSTer and a few new Arcade cores (e.g. Ghost N Goblins and 1942) appeared first on MiST recently.

 

MiSTer is kind of a successor though. What prompted MiSTer development was the lack of HDMI amd the bigger FPGA available for cheap via the DE10 board. It's true that it's subsidized hardware, but there are other boards that can receive the MiSTer framework (with some minor porting work, somebody ported some cores to the DE0 SoC board).

 

I was an early MiST adoptwr (from 2015) and I have to say it's amazing to see how much has progressed.

[youxia]

I also thought MiST was discontinued...sometimes it's hard to keep up with Lotharek's projects

Had a quick look and it seems it currently sells for about 230$. Seems strange seeing how it offers less than MiSTer which can be got for less. Am I missing something here?

[Newsdee]

I also thought MiST was discontinued...sometimes it's hard to keep up with Lotharek's projects

Had a quick look and it seems it currently sells for about 230$. Seems strange seeing how it offers less than MiSTer which can be got for less. Am I missing something here?

I have no idea what their margin is, but hardware MiST clones now exist at only a slightly smaller price (e.g. MiSTica) so I suppose they are sold quite close to cost.

 

MiSTer uses a broadly available FPGA training board targeted at education, and as such has much higher production volumes. It is cheaper if you don't count for the case.

 

The thing is that the Cyclone III of the MiST can be manually soldered on a simple two-layer PCB, so the MiST is (relatively) easier to clone from scratch if needed. In contrast, the Cyclone V used by MiSTer is much more expensive per unit and needs special equipment for soldering BGA, and a multi layer PCB (ideally) to route all connections around. So there is not much incentive to make a homebrew MiSTer, when getting a board from Terasic is so easy.

 

Also one should bear in mind that when the MiST came out these training FPGA boards costed much more (300 to 500 USD).

Edited February 27, 2019 by Newsdee

[youxia]

MiSTer uses a broadly available FPGA training board targeted at education, and as such has much higher production volumes. It is cheaper if you don't count for the case.

 

Yeah, I can see it's cheaper, substantially - plus it offers more, hence my question. I'm sure all the intricacies you describe are correct, it's just to a potential buyer such as myself these are not really relevant - if something is subsidized or not, or whatever the profit margins are.

 

When you mentioned that MiST is still on sale, I went to check it on a whim because I thought it may be now heavily discounted. I don't need HDMI and my main interest is old computers, so could maybe get this instead of MiSTer. But with this price...

 

Basically, I'm just curious why would anybody buy MiST when there's MiSTer?

[RodLightning]

My two cents, MiSTer is still in the early stages of Atari ST support. End users who aren't developers will have to wait for that to evolve. MiSTer/DE10 nano has a mini usb and whatever ports your i/o board (if you buy and add one) has on it. Fully populated boards can offer analog vga style port, 3.5mm stereo jack, optical audio and usb type jack that is actually not usb, but serial; and midi in/out can eventually be built on that port for expansion. There is a sdram board required for some cores that is also optional, but nice to have. The nicer i/o board has sdram built in. Adding at least a 4 port OTG type hub is also pretty much required for mouse, keyboard, game controllers etc. There are some GPIO ports similar to raspberry pi, usable by developers if they so choose.

 

On the other hand, a true MiST box offers STe core, real midi ports and analog video out for use with CRT. It's ready to go out of the box.

 

MiSTer has more potential, but currently requires more effort to get up and running. I have a MiSTer box built at present and like it very much. I have never owned a MiST.

 

MiSTer will keep you reading through forums and cruising github if you want to have the absolutely latest cores and utilities running, but it's a lot of fun.

[Newsdee]

MiSTer will keep you reading through forums and cruising github if you want to have the absolutely latest cores and utilities running

No longer... check out the updater script, once in the SD card you can run it from OSD and it will update all binaries if you're connected to your network.

[RodLightning]

I have been manually messing around with midilink 2.0 and 2.3 for MiSTer lately, trying to see what is presently possible. Success getting the 486 core playing midi songs. Not a lot of joy so far with midi on fxcast atari st core, except with tcp/ telnet out to bbs which is working great.

 

Keeping in the spirit of this particular thread though, I think the excellent MiSTer arcade machine cores are worth it alone in buying the de10 nano hardware.

 

And yeah, I do want to get that updater script installed and get everything onto one sd card. I currently have two cards, one with the computer cores and another for game consoles and arcade systems. At the rate the project is expanding, I think a 128gb sd card will be needed soon, if you account for software, hdd images and media used by the various systems.

[derFunkenstein]

SmokeMonster has been an incredible advocate for the setup. Im seriously considering getting one but Im content to watch until NeoGeo is documented and the 16-but systems are all good. From what Ive read the Genesis needs some love still.

[Newsdee]

Great news on the HDMI front, there is now a zero latency mode:

https://www.retrorgb.com/mister-hdmi-core-now-sub-1-frame-of-lag.html

[youxia]

Humans can be rather clever:

 

[SegaSnatcher]

about 14K LE:  
- Collectorvision Phoenix, Colecovision + more, Open Source cores (Colecovision, Atari 2600), and possible Commercial cores in the future.

about 49K LE: 
- Analogue Super NT, SNES only, commercial core
- Analogue Mega SG, Genesis/Megadrive + more, commercial cores

about 110K LE:
- MiSTer project, Arcade, Classic Computer and Console cores, Open Source cores.

Edited August 1, 2019 by SegaSnatcher

[Keatah]

I overheard about a kit for MiSTer, it includes some kind of PCB kit and other stuff. It's an add-on package. Anyone know what I mean?

[SegaSnatcher]

3 minutes ago, Keatah said:

I overheard about a kit for MiSTer, it includes some kind of PCB kit and other stuff. It's an add-on package. Anyone know what I mean?

There is add on board that adds analog out and a fan.  They call that the i/o board. 

Is this what you mean?  

http://misteraddons.com/product/mister-io-board-xl-v5-6-with-heatsink-and-fan/

[SegaSnatcher]

If the mods allow you to edit your OP, I was thinking perhaps change thread title to "General FPGA Discussion" or something similar?    

Really would like your thread to be the new go to for all things FPGA and that way Kev's thread can be focused to his personal projects.

 

[Newsdee]

1 hour ago, Keatah said:

I overheard about a kit for MiSTer, it includes some kind of PCB kit and other stuff. It's an add-on package. Anyone know what I mean?

MiSTer is actually a PCB kit more than anything else. It was built as add-ons to a commercial board (DE10 Nano) supporting a OSS framework, but it is not exclusive to it.

 

At least a couple of persons ported the framework and a habdful of cores to other boards.(eg DE0). Those boards are pricier or have other caveats (smaller FPGA, no HDMI, etc) so they remain niche.

 

For the layman these are irrelevant, it's easier to tell people to get the cheaper board, but that also means MiSTer (or a descendant project) will outlive the DE10 supply drying up.

 

If that ever happens, because that board is popular with university courses.

 

Edited August 1, 2019 by Newsdee