A second PS1 core coming

[jca]

[caad]

Worth noting that it is *not* for the MiSTer but runs independently on the DE-10 Nano (in another environment).

So it is not a direct competitor to the other core, and it makes different tradeoffs.

[jca]

The thread is somewhat confusing. See these 2 posts:
!!! IMPORTANT !!!
This is using the SAME HARDWARE as MiSTer but using a different framework. So it is compatible with any MiSTer hardware but not using the feature of MiSTer framework. Sorry for the ambiguity.
______________________________________________________________________________________________________________________________
So later on, if people want to help port the core over MiSTer (or when I find time to do it myself) then it will work over MiSTer. For now, it is only the same HARDWARE. Sorry for the confusion.

I guess it can be run on MISTer using a different SD card using his own framework.

[sonik]

It can be loaded from mister?
Or it needs an different way to load? Boot from another card?

[hiddenbyleaves]

A bit redundant now Rob is on the case.

[N.Master]

It can be loaded from mister?
Or it needs an different way to load? Boot from another card?

it is a PSX emmulator for DE-10 Nano board
not a Mister core

Same hardware, different software . Like a PC running Linux and the same PC running Windows.

[Milspex]

its like Scummm basically

[justaguy]

I guess it can be run on MISTer using a different SD card using his own framework.

FPGAzumSpass had mentioned in a Patreon update that something like 17% of the FPGA is devoted to the MiSTer framework in his PlayStation core. While he's obviously making it work for PSX, it got me wondering what might be achieved with a stripped-down, barebones framework that wouldn't be possible otherwise. (Not to say that the MiSTer framework suffers from feature bloat or anything--quite the opposite.)

[Neocaron]

Very nice to see, they probably are very different beast in term of how they do things. Glad Lexar was able to go back to it without any pressure.

[MikeJ]

A bit redundant now Rob is on the case.

Laxer's core was in dev for years. It was parked in May at ~95% done.
Laxer reached out to Robert to see if he wanted to collaborate, but Robert decided (and I do get this) to start his own.

All good.
I've been helping Laxer on and off for years and we decide to bring it up on the DE just to get it over the line and completed.
I asked Ash if he could run it up on the mister framework, but that's proven difficult so far - but I'm sure it will happen.

We use a different SOC setup, including the AXI connections and a different linux environment - we are not doing things on our own just for the hell of it
/Mike

[Waifu4Life]

Wonder why he decided to take this route?

[sonik]

it is a PSX emmulator for DE-10 Nano board
not a Mister core

Same hardware, different software . Like a PC running Linux and the same PC running Windows.

I know.
But how to use it on a "mister hardware". Mister is not just the de-10 nano There's the pins used for the memory and I/O...
Would booting it from a SD card work? I don't think so.

[TLPD-AVW]

its like Scummm basically

This is a bad analogy, as ScummVM does not utilize DE-10 Nano's FPGA resources. This core does.

[MikeJ]

I know.
But how to use it on a "mister hardware". Mister is not just the de-10 nano There's the pins used for the memory and I/O...
Would booting it from a SD card work? I don't think so.

We've kept compatibility with mister SDRAM and the IO module (as this is what people have).
You just need to flash an sdcard with the image and swap.

[sonik]

Nice!
Really interesting to know. Thanks for clarifying.

[bbond007]

its like Scummm basically

It's currently a core for an alternate framework compatible with the DE10-Nano and MiSTer related hardware expansions.

From the Laxer3A's ReleaseMemo.TXT:

"I hope in the future that once the core is more stable, it could be integrated / picked up by
somebody interested to port it to MiSTer but for now, given my limited amount of time on the project
I prefer to focus on core stuff. That may change in the future, may be I could do it myself,
I don't know yet and haven't decided a thing about that."

This is a bad analogy, as ScummVM does not utilize DE-10 Nano's FPGA resources. This core does.

Yes, however, ScummVM does use the FPGA for framebuffer & audio via the menu.rbf core.

[aberu]

Wonder why he decided to take this route?

Because Laxer and Mike have been good friends for awhile and Mike is using the DE10-Nano as a springboard to launch his new framework. MikeJ's actual target hardware is different than the DE10-Nano, however his current target hardware has massive supply issues everywhere and 48+ week lead times last I checked. He's probably just wanting to get the framework going and fleshed out so he can eventually launch the real target.

A bit redundant now Rob is on the case.

Laxer's core was in dev for years. It was parked in May at ~95% done.
Laxer reached out to Robert to see if he wanted to collaborate, but Robert decided (and I do get this) to start his own.

All good.
I've been helping Laxer on and off for years and we decide to bring it up on the DE just to get it over the line and completed.
I asked Ash if he could run it up on the mister framework, but that's proven difficult so far - but I'm sure it will happen.

We use a different SOC setup, including the AXI connections and a different linux environment - we are not doing things on our own just for the hell of it
/Mike

Hey Mike, hope all is well.

I thought it was 95% done in his simulation model (done in terms of the known architecture being described, not in terms of emulation compatibility metrics). The integration was not at 95% done though. This is where the integration side that ultraembedded worked on a lot was left:

https://github.com/PS-FPGA/ps-fpga

Hardware
CPU
GPU
GTE
SPU (stub)
MDEC (stub - produces a white screen)
DMA
CDROM (SPI / RAM based 'CDROM')
Various peripherals (timers, interrupts, serial ports).
DVI - driving DVI/HDMI directly from the FPGA pins (TMDS)
Controller - can be hooked up to a standard controller and memory card.

Limitations
Currently running at 30MHz instead of 33.8MHz (for DVI purposes).
CDFW needs completion.
Missing SPU / MDEC (lack of functioning SPU causes issues for a number of games).

The SPU can be one of the most difficult things to integrate, and the integration didn't even have the MDEC yet which is a major component. So I do not believe it's accurate to imply the core was 95% done back in May.

The integration was also for an Artix Arty-7 A100T board, that's why it's taken quite a while to port it over to the DE10-Nano and over to your framework.

I look forward to the sources being posted so I can see all the stuff that was added, how it differs from Robert's core (which is fully open sourced right now). If binaries are being distributed now, hopefully the source is being prepared to be distributed soon. I know that the project was a dual license (GPLv2 and Commercial). I may be ignorant, but I think GPLv2, regardless of dual licensing structure, obligates full source code releases upon binary releases. However, since this is a preview/alpha it's probably not that important now.

Does the current core being distributed still use this source? --> https://github.com/ultraembedded/core_u ... 8d8a24b713 - If so that is combining Apache 2.0 licensing with GPLv2 and a commercial license. GPLv2 supposedly has incompatibilities with Apache 2.0, but I don't think they are relevant to the content of this project.

Does the current core still use ultraembedded's core axi cache? --> https://github.com/ultraembedded/core_a ... e4d35ad0be - I would assume not since you are using the AXI from Intel on the DE10-Nano, but just wondering because this also uses a BSD license which just adds to the licensing complexity.

I'm looking forward to seeing the source so others could see how all of these significant challenges you have been up against were solved.

[Blitzwing]

Wonder why he decided to take this route?

I wouldn't want to speak for him, but my understanding that I get is that he has his only particular kind of workflow and in particular doesn't pick up already commenced work... basically to get a firm grasp of all aspects of the target and the code. Also allows to potentially incorporate features from other cores like savestates, from the start.

[softtest9]

Very pleased to see Laxer3a back in the game. Amazing work! Whichever core ends up better overall, this is very positive. If we look at the emulation world, we can see that all of the different "competing" projects have ended up benefiting each other, whether it's through borrowing code, making use of other developers' documentation, or just sharing ideas.

I can also understand the desire to work on one's own project even when there is another project doing something similar. Adopting someone else's work and molding it into what you want can be a big challenge in and of itself, even though it might seem like the obvious way to go at first glance.

I'm also curious about Mike's board and framework. Are there any details about what the board will be like (logic elements, features etc) or how the framework differs from the MiSTer? Anyway, can't wait to try this out. I got a second SD card ready whenever there is an official release.

[Caldor]

I made an image of the SD card yesterday, updated the core and such. But there are some USB compatibility issues. A few do seem to have USB working, but many cannot get it work, not with the MiSTer USB hub, or even a direct OTG USB adapter. MikeJ said something about an update coming later today.

But I also read it might still be possible to get it to boot using different bioses. It seems to use the scp1000 bios by default, but for some games it might help to use the scp1001 and rename it to 1000. So a game might run without USB input because you can setup the ini file and it will boot directly into the game. For now you have to convert the game you want to play, and I think it should even support 700MB CDs. Until recently it was loading CDs into the DDR RAM I think.

Last weekend I was testing Peips core, so this weekend I plan on testing Laxers core, and MikeJs framework The plan for MikeJs framework from what I hear is to make a more high end FPG that might be able to handle higher end cores, or maybe just do more with them. Currently there is room for improvement with the Amiga and 486 core, and we will see how PSX turns out, but there might come other cores where extra FPGA power will help make it more compatible or add more advanced features.

[MikeJ]

The new board is based on Xilinx Ultrascale+ MPSoC. I've been working directly with Xilinx for some time. Layout was near complete with the 3EG, but it made sense to switch to the Kria module when that was released. I'll show where we are shortly.

Aberu, the MDEC was complete and the SPU mostly in May. They were not committed to the original ps-fpga repos.

Once we have the SPU and some bug fixes in place, hopefully in a few days, I'll have time to clean and release the new ps-fpga repos.
The move from the A100T was not a big deal, what has taken time is getting our firmware to a state where we could run the AXI bridges etc.

Caldor, the controller support is external to the core. We are using SDL2 and it's super fresh - it was integrated a few days ago. I'm quite impressed how much is working already.

[softtest9]

I just looked up Xilinx Ultrascale+ MPSoC and it seems insanely powerful. I can't wait to see how all this develops!

[metal_slugger]

Will there ever be (or is there already) a way to potentially "dual boot" the DE10 into either the MiSTer framework or any other framework, from the same storage device? Failing that, swapping out an SD card for a different framework is no problem anyway.

I guess this may be a redundant question as the MiSTer framework is not just the SD card contents but the code / integration of the I/O board etc too....

[Jegriva]

Thank you all for the clarifications!

[MikeJ]

Will there ever be (or is there already) a way to potentially "dual boot" the DE10 into either the MiSTer framework or any other framework, from the same storage device? Failing that, swapping out an SD card for a different framework is no problem anyway.

I guess this may be a redundant question as the MiSTer framework is not just the SD card contents but the code / integration of the I/O board etc too....

We kept the same I/O board pinout and SDRAM.
Originally we were going to make our own, but that would fragment the market and it's better we focus on our own hardware.

The main reason we had to do our own boot/rootfs was so we had control over the SOC configuration. This is all documented now and builds with github actions. We are using barebox rather than uboot as it's a bit more modern, so I don't think a dual configuration is possible.

I appreciate we are behind mister on features and core support so it's not realistic to expect people just to use our stuff.

Can the de10-nano boot from usb? If so that might be an option - open to suggestions.
/Mike

[softtest9]

You can load cores from USB, just not Main_MiSTer. Could one expand on this and make the SD card a "dumb" bootloader, that simply loads everything from USB? This would also help MiSTer users limit the strain on the SD card slot (no more taking it out just to update Main_Mister), if they got a USB drive. Not sure if this would be helpful for dual-boot though.

[aberu]

I just looked up Xilinx Ultrascale+ MPSoC and it seems insanely powerful. I can't wait to see how all this develops!

Did you look up the price?

2021-12-01_9-51-03.png (72.12 KiB) Viewed 20384 times

2021-12-01_9-54-20.png (133.72 KiB) Viewed 20384 times

The SoM that he is targeting, the Kria K26 SoM, is exceedingly expensive now and is having supply issues. Not only that, this module will require a pretty complicated carrier board to be designed by MikeJ and would likely need to be a 6 layer board (maybe he can do it with 4 layers), so manufacturing would not be cheap. All in all, if the pricing situation doesn't change for the Kria anytime soon I expect his price point to be minimum $600-800 range. If it is a premium system, that would be awesome. I'm just trying to help people not go out and buy something too fast, definitely take your time as a consumer and do your research when things get announced.

It's an absolutely awesome hardware platform with very high potential, but take your time reading up on it, and watch how things play out since none of us can know all the details yet or how supplies will be or what the price will be like, or even if the module itself may have unforeseen problems for this kind of application (I don't expect it will).

[Atohmdiy]

Last time i check the Cyclone V alone is also a more expensive than the de10-nano. Problem is Terasic did not release the de10 for the retro gaming, but as general dev board for education. I don't think a board specifically designed for retro gaming with much less audience could be discounted like that.

So a board with this xilinx will be more expensive, no doubt, but the questions is more for what purpose. Ps1 and saturn for the mister seems to be in good shape for a release relatively soon. There is discussion for the feasibility of the N64. Next would be probably the Dreamcast, surely impossible ?
Here there seems to be a gpu with some capabilities, and a more powerfull cpu. So the goal is to offer a new plateform that could go beyond what the de10 can do with some sort of hybrid emulation ?

Would be great for sure, but is it really doable ? I mean if you want enthusiast to buy an expensive and powerful new platform you need something more to offer, like core for more advanced system. These core are surely doable in term of hardware with this new thing, but how many time and resources to develop them ? The ps1 core from laxer took years and provoke a burnout of his dev... so i imagine a N64, dreamcast or ps2 core...

So you will offer something a lot more expensive, that would offer nothing new to the community probably for years. It's risky, considering you are building a custom hardware... in that situation it is probably wiser to wait for a general purpose and commonly available dev board from a big company ?
Sorge did it just right when he choose the de10-nano and it's probably the major reason for the success of the mister, instead of something custom.

Last thing in the "what's the point" argument, with more advanced system you enter the real of 3d, and software emulator can do a lot of amazing thing to increase the graphics. Not sure what this mali gpu could do, but i am sure it's not what my 6900 xt can do If i want to play a 3d ps2 game, i run it with pcx2 with my oled tv with all the fancy option. If one day a ps2 core exist for this new platform, i am pretty sure the mali gpu couldn't do what pcsx2 is doing, so 99% of people would just use pcsx2 with their modern display. I am with the 1% left that would enjoy playing with my crt, and surely this pateform would give the same benefit of the mister : easy setup with a crt (retroarch is giving me headache with my crt...), everything cleaner and more embedded with less lag etc...

Anyway, it's great news for the ps1 core from laxer. Would have been a shame so much work and tears for just abandon it so close to the end.

[Caldor]

Will there ever be (or is there already) a way to potentially "dual boot" the DE10 into either the MiSTer framework or any other framework, from the same storage device? Failing that, swapping out an SD card for a different framework is no problem anyway.

I guess this may be a redundant question as the MiSTer framework is not just the SD card contents but the code / integration of the I/O board etc too....

We kept the same I/O board pinout and SDRAM.
Originally we were going to make our own, but that would fragment the market and it's better we focus on our own hardware.

The main reason we had to do our own boot/rootfs was so we had control over the SOC configuration. This is all documented now and builds with github actions. We are using barebox rather than uboot as it's a bit more modern, so I don't think a dual configuration is possible.

I appreciate we are behind mister on features and core support so it's not realistic to expect people just to use our stuff.

Can the de10-nano boot from usb? If so that might be an option - open to suggestions.
/Mike

I think it can. At least when you boot on the MiSTer you can switch to booting from USB instead of the SD card. I boot from the SD card though as for some reason I cannot run updates to the USB, it seems to consider everything write protected. But its not a problem when I am running the Minimig or AO486 core f.ex. that do write to virtual hard drives and floppy disks. Might just be something I am doing wrong.

But I got the Replay framework running and got USB working. The problem seems to have been that the power on my MiSTer setup goes through the USB hub and then into the MiSTer because the USB hub has a power switch and more. The Blisster hub. When I put the power directly into my DE10-Nano I could use my keyboard without issue, connected directly to the USB hub micro USB output on the DE10 Nano with a OTG USB adapter.

So I have been testing Quake 2 and Diablo on it. I will make some videos this weekend I think, testing some more games.

[softtest9]

All in all, if the pricing situation doesn't change for the Kria anytime soon I expect his price point to be minimum $600-800 range. If it is a premium system, that would be awesome. I'm just trying to help people not go out and buy something too fast, definitely take your time as a consumer and do your research when things get announced.

5-600 000 logic elements, did I get that right? With an FPGA like that, I would not be surprised if we got Dreamcast, Nintendo DS, GP32 and a PC core with Pentium II + 3DFX (maybe even Windows XP would be usable). All of those systems are pretty annoying to mess with today, especially bulky old PCs that require IDE drives and so on. So yeah, I will see how things play out, but the price isn't going to stop me if we start seeing these kinds of cores.

Last thing in the "what's the point" argument, with more advanced system you enter the real of 3d, and software emulator can do a lot of amazing thing to increase the graphics. Not sure what this mali gpu could do, but i am sure it's not what my 6900 xt can do If i want to play a 3d ps2 game, i run it with pcx2 with my oled tv with all the fancy option. If one day a ps2 core exist for this new platform, i am pretty sure the mali gpu couldn't do what pcsx2 is doing, so 99% of people would just use pcsx2 with their modern display. I am with the 1% left that would enjoy playing with my crt, and surely this pateform would give the same benefit of the mister : easy setup with a crt (retroarch is giving me headache with my crt...), everything cleaner and more embedded with less lag etc...

First of all, the GPU is almost never the bottleneck with emulation. PS2 and beyond are problematic because of high CPU requirements. None of these modern emulators can get full speed without a dynamic recompiler. So that's actually another area where an FPGA might really beat software emulation. All of your other arguments about the PS2 can also apply to the PS1, Saturn and N64. You are of course free to use what you want, but personally I'm more interested in accuracy and low latency, than any kind of glitchy enhancements. That's why I got the MiSTer in the first place.