Commander X16

On 10/11/2021 at 7:22 PM, Scott Robison said:

I can't "defend" (nor can I "attack") how project leadership is handling communication with the community. For me personally, I look at it as "I would like to know more, as much as possible, but no one owes me more info; in like fashion, I do not 'owe' the project my allegiance". What each of us has to decide for ourselves at this point is "we know how this project is working; when do I reach the point that staying is no longer of value to me".

In my messages I have deliberately made an effort to avoid any suggestion that the project owes or should do certain things for us. I am merely offering my thoughts on what I subjectively think would be in their best interest, i.e. an occasional comment on this thread given the gravity of the topic and the amount of requested feedback gathered. It would simply seem like a good idea for them IMO.

I have been around on the Internet too and I know how it can be at its worst. However, what I see here is actually a very reasonable-looking community by those standards. I think the occasional comment to the requested feedback or a simple note on continued planning would do more good than harm for the project ands its owners. Obviously they have judged (deliberately or not) otherwise, but just my thoughts.

On 10/11/2021 at 1:34 PM, Janne Sirén said:

In my messages I have deliberately made an effort to avoid any suggestion that the project owes or should do certain things for us. I am merely offering my thoughts on what I subjectively think would be in their best interest, i.e. an occasional comment on this thread given the gravity of the topic and the amount of requested feedback gathered. It would simply seem like a good idea for them IMO.

I agree you've been deliberate, and my comments were not directed any any particular person, just a general observation. Sorry if they came across differently, that was not my intent.

Thanks for the clarification, no worries of course!

I think that the best way to implement the commander x 16 e should be an fpga with 512 kB of block ram

Ram could be divided as follows 64 base + 128 video +320 banked

inside such a capable fpga all the chipset should find room.

 

On 10/11/2021 at 5:44 PM, Fabio said:

I think that the best way to implement the commander x 16 e should be an fpga with 512 kB of block ram



Ram could be divided as follows 64 base + 128 video +320 banked



inside such a capable fpga all the chipset should find room.

How much does an FPGA with 512KB of block ram go for?

On 10/11/2021 at 2:06 PM, picosecond said:

Here is an over-simplified but still useful way to visualize FPGAs.



Imagine a huge Ben Eater style breadboard prepopulated with thousands of simple TTL gates and flip-flops, but no wires.  By itself this logic does nothing, but by adding the right wires one could implement many possible useful circuits.



Programming an FPGA is analogous to plugging wires into this breadboard.

Quite, which is why I rather refer the process of implementing the routing as "wiring" the FPGA, rather than "programming" the FPGA, which causes all manner of confusion to people who know a bit about programming and very little about CPLDs and FPGAs.

This recent discussion has raised a question in my head: How much would it cost to turn the VHDL for VERA into custom-fabricated fixed silicon?

On 10/11/2021 at 10:21 PM, kelli217 said:

This recent discussion has raised a question in my head: How much would it cost to turn the VHDL for VERA into custom-fabricated fixed silicon?

That's a difficult question to answer exactly, but from what I can find via Google searches, you're looking at an up front cost of multiple tens of thousands of dollars, then a volume of at least 10k per year to bring the costs down to the sub $2 range per IC. It's not impossible but not practical for the expected scale of something like the X16. That all assumes some of the least expensive processes would be usable.

There is also the point that once you go fixed--silicon, that's it. So, it kind of makes sense it is the last step in a project like this – If it is to be done at all. Even if the feature-set would be fixed early on, the upgradeability of an FPGA can be very helpful in low-volume hobbyist projects. We do have some precedence for this though in the Commodore community: the hobbyist C-One (FPGA) evolved into the higher volume C64 Direct-to-TV (ASIC).

On 10/12/2021 at 4:33 AM, Ju+Te said:

I'm not experienced with FPGA at all, but it sounded to me like they could be programmed that they form some kind of PCB of some generic (TTL) parts. If I understood this correctly, the schematic already exists (as "program" for the FPGA). I might be completely wrong, so please correct me.

You are correct, and what you suggest is a good analogy. There is a circuit out there as a program to be simulated on a computer known as a FPGA.

A FPGA is a computer itself, it has computing hardware, runs a program, takes input gives output same as any computer.

At the moment, the entire experience of the X16 is 'available' in the simulator, if that is what you desire. FPGA is another simulator which can run the same simulation. The simulation can be ported to windows, linux, or FPGA as you please.

Personally, I don't see it as the same as the 'real thing' where you can hold and understand each part and with that understanding you get the best abilities to program, repair, expand, adapt the design to your desires. Just as to understand the simulator on Linux, you have to understand the software running the simulation, the operating system supporting the software, the hardware supporting the operating system, to get the most out of it. You have to learn the software and toolchain and hardware which supports the FPGA itself in order to get the most out of the simulation. It is what I consider to be an unnecessary layer of abstraction from the circuit itself.

Perhaps the simulated circuit is indeed so sprawling and verbose that economy prohibits discreet parts, however I think the circuit objectives are modest. With todays lower costs of all the parts needed to build 80's level computers, I see the actual real world as a good environment for the computer circuit to exist in discreet components.

I see putting in the effort to optimize a circuit so it can be realized in discreet parts with low chip-count as necessary for the desired experience of the end user because how can you interface with a simulation? There are the added obstacles of learning every layer of abstraction, and in the end, you'd just stick to learning one or two which most suit your goals. Learning programming on the linux machine rather than learning programming on the linux machine PLUS learning programming in the simulation so you can modify and interface to the simulated x16 better and create addons, well there is the extra layer of busy work right there.

The Pi is bad enough as you cannot solder every signal and even where people do not want to pick up an iron there are benefits and innovation from other people with that same hardware to benefit from.

Less layers of abstraction equates to less learning required for access. So I'm personally against FPGA, it's not worth learning for me, I'd sooner just design AND THEN OPTIMIZE the circuit so its easily built economically.

On 10/12/2021 at 5:06 AM, picosecond said:

Imagine a huge Ben Eater style breadboard prepopulated with thousands of simple TTL gates and flip-flops, but no wires.  By itself this logic does nothing, but by adding the right wires one could implement many possible useful circuits.

Has anyone here built his video card circuit? I have not, but I think that some of the inverters can be omitted from his circuit without changing it's function or even the circuit beyond the omissions. I should try to contact him perhaps, but I'm sure he's busy, maybe someone else has built it.