Are you keeping up with the Commodore?

[Pixel_Outlaw]

Because the Commodore (really has been) keeping up with you.

Here is a shot of the assembled prototype for the new Commodore 64.
HDMI output, Cherry mechanical keyswitches, DVD reader w optional blueray player. Can run modern OS like Windows.

http://www.commodoreusa.net/CUSA_C64Prototype.aspx

[Lance Boyle]

Commodore USA, LLC was founded by Barry Altman in April of 2010, with the express purpose of reviving and re-establishing the famous Commodore computer brand. We are Commodore and AMIGA fanatics, just like many of you. We ask ourselves what could have been, and we are appalled by Apple revisionism. Commodore is back, and we're determined to bring the much loved brand back to the mainstream and restore its prominence in the tech industry to that which it richly deserves. It ain't over 'till we say so.

GOOD LUCK WITH THAT.

[Pixel_Outlaw]

It is definatly a niche genre for the C64, however they do have other less retro looking products.

I for one will purchase the new C64 just to tinker around in retroville plus the keyboard should feel great since they have Cherry keys inside. I still like the idea of a self contained keyboard computer.

[Specineff]

If they can make something like a consolized Amiga that can run all games (Image or actual floppies), I'm SO there.

[drauch]

If they can make something like a consolized Amiga that can run all games (Image or actual floppies), I'm SO there.

THIS.

[dcharlie]

Guys,
For the "consolised Amiga" - Welcome to the MiniMig.

Runs images, takes traditional joystick/mouse, attach a keyboard, output via VGA.



Waiting for a FPGA C64 in a similar form. Massive C64 fanboy here

[flux]

I don't get it. So it's just basically a laptop shoved into a C64 replica case? Running a C64 emulator? That seems really pointless and like it's for people who mistake nostalgia for quality. Especially since you can get an actual C64 for pretty cheap on eBay if you actual want it to use it for the hardware.

edit:
Now, if someone were to make a perfect replica of the SID and, say, put it in a box with some knobs and a MIDI in, that would be awesome. Even that has a pretty good emulator and a ton of soft synths replicating it.

[null1024]

The lame bit is that the machines don't even run real AmigaOS 4, they're just PC machines. Shameful.

[Herr Schatten]

Now, if someone were to make a perfect replica of the SID and, say, put it in a box with some knobs and a MIDI in, that would be awesome.

Already done.

[Ex-Cyber]

Waiting for a FPGA C64 in a similar form.

What about the C-One?

[dcharlie]

considering a C-one but would prefer to get a fully preconfigured unit.

[kernow]

Or a real C64 with SD card loading cartridge.

[djvinc]

oh my god, the Commodore prototype looks so lame inside.
This is not what custom hardware is about : when you open a modern console for instance, you find custom PCBs, chips, connectors etc.
Damn fanboys, don't you want to learn *real* hardware design ??

[louisg]

That Minimig looks awesome. Any word on compatibility, and does it set a VGA monitor to 50hz or what?

[spadgy]

I'm hugely nostalgic about Amiga and Commodore - I'll admit that. So this does set my heart aflutter.

I'll admit, it looks like they could have been more ambitious with their design and implementation of that C64, but I'm certainly curious about those Amiga 1000, 2000 and 3000s.

[Ed Oscuro]

Waiting for a FPGA C64 in a similar form.

FGPA's only benefit is it's cheap...it isn't guaranteed to work like the real thing. Reasons being the FGPA has to simulate the functions of the original chip, and might run at a different or imprecise clock rate. Maybe it's a step up from some other designs, but in this day and age I think it should be possible to find a small fab somewhere and make a run of both the original 6581 and one of the later revisions, to be switchable on the outside of the case. In truth these new stabs at the design probably aren't much cheaper to design than the original, and doing it "right" would probably be too expensive, but I think that's how it should be done.

[Ex-Cyber]

FGPA's only benefit is it's cheap...it isn't guaranteed to work like the real thing

Most FPGAs are also reconfigurable. If you get something like a C-One, you can make it be a C=64 for a while, and then reboot and load a set of Spectrum cores, then later reboot and run your own original design. It's pretty neat.

Reasons being the FGPA has to simulate the functions of the original chip

Saying that the FPGA "simulates" the functions of the original chip really gives the wrong idea. It's not a simulation any more than a re-design for a modern fab process would be a simulation. In the case of something like SID, analog filters would probably be replaced by digital ones, but that would probably be done in a re-design too, since the original filters are infamous for being quirky and out-of-spec.

in this day and age I think it should be possible to find a small fab somewhere and make a run of both the original 6581 and one of the later revisions

Even if you could find or reconstruct the original layout (not impossible, but probably difficult/expensive), I think you'd have trouble finding a fab that's both willing and able to manufacture it. IIRC it was laid out with a specific MOS Technologies 7-micron NMOS process in mind, a process that was already virtually obsolete at the time. And it's a mixed-signal chip, to boot.

[Ed Oscuro]

FGPA's only benefit is it's cheap...it isn't guaranteed to work like the real thing

Most FPGAs are also reconfigurable. If you get something like a C-One, you can make it be a C=64 for a while, and then reboot and load a set of Spectrum cores, then later reboot and run your own original design. It's pretty neat.

That really doesn't address what I'm talking about, though you're right, I'm dismissing FPGA too harshly there. Yes, you can make a FPGA do something very well, or half-ass it. I'm not saying that the FPGA implementation is going to be wildly off, but I wouldn't hold my breath that it will be. Anyway, aren't pretty much all modern FPGAs reprogrammable anyway? Otherwise you'd have a PROM, which is also field-programmable but just once (usually). And who the hell wants to wait to reconfigure a FPGA in actual use? It would be better to either use the sizable space in a modern FPGA to replicate both at once.

Reasons being the FGPA has to simulate the functions of the original chip

Saying that the FPGA "simulates" the functions of the original chip really gives the wrong idea. It's not a simulation any more than a re-design for a modern fab process would be a simulation. In the case of something like SID, analog filters would probably be replaced by digital ones, but that would probably be done in a re-design too, since the original filters are infamous for being quirky and out-of-spec.

I also worried about using the term for the reason you mention, but you get what the remaining problem is. I should say "replicate" perhaps. Even if you duplicate the internal functions of a chip, it's not the same as the actual chip itself. This may not create any obvious differences to a person, but I wonder if anybody has worked from original tapes or how much detail the (now expired) patents provide. And since the chip went through multiple revisions, I doubt the original patent would provide an accurate-enough picture of what the chip actually worked like.

in this day and age I think it should be possible to find a small fab somewhere and make a run of both the original 6581 and one of the later revisions

Even if you could find or reconstruct the original layout (not impossible, but probably difficult/expensive), I think you'd have trouble finding a fab that's both willing and able to manufacture it. IIRC it was laid out with a specific MOS Technologies 7-micron NMOS process in mind, a process that was already virtually obsolete at the time. And it's a mixed-signal chip, to boot.

I did say it would be expensive, right? If you couldn't find the original tape layout or if the patent didn't provide enough data about the actual workings (as mentioned before) you could take an example of the chip to be torn apart, rather like how MAME's decapping project works, except here doing it with the aim of getting an idea of the entire chip's layout.

Anyway just some random thoughts...but I've always been suspicious of the accuracy of mainly commercial projects in duplicating 100% accuracy with an original architecture, kind of like how some FPGA implementations of the original PC architecture are just called "compatible," not "exact" (not that in practice you'd need it for 99% of PC software being developed for clones, but for a single-source architecture like the C64 it might matter), or how popular Intel chips were cloned by competitors, and so on.

[Ex-Cyber]

Anyway, aren't pretty much all modern FPGAs reprogrammable anyway?

The vast majority are. The exceptions are mostly intended for military/aerospace applications and use antifuse programming.

And who the hell wants to wait to reconfigure a FPGA in actual use? It would be better to either use the sizable space in a modern FPGA to replicate both at once.

That would be stupid in a volatile FPGA (the most popular and usually the most cost-effective type), since it has to be configured at power-on anyway. Modern FPGAs don't necessarily require users to plug in a JTAG cable; they have several configuration modes, and it's not too hard to have e.g. a microcontroller load the bitstream from an SD card.

I did say it would be expensive, right? If you couldn't find the original tape layout or if the patent didn't provide enough data about the actual workings (as mentioned before) you could take an example of the chip to be torn apart, rather like how MAME's decapping project works, except here doing it with the aim of getting an idea of the entire chip's layout.

What I'm getting at is that especially with chips that old, the layout isn't necessarily enough, particularly for a mixed-signal chip. You don't know what ended up being resized to compensate for manufacturing quirks. Actually you also have the reverse in SID's case - they didn't have time to fix all of the manufacturing quirks. The design was done on an insane schedule, and some major corners had to be cut to produce a chip that worked at all. So the actual chip behavior doesn't directly follow from the intended design/layout. This would be no big deal if the quirky behaviors were universally regarded as bugs and avoided, but programmers and composers actually took advantage of at least one of them.

[Ed Oscuro]

Agreed on those points - last one in particular seems to point to reverse-engineering the chip by tearing one apart. Of course, with the FPGA option I suppose you could program not just the original SID functionality but add sliders or toggles as well, i.e. the "buggy" 12th bit that caused distortion - in any case, the proof will be in how well this current implementation works...won't likely be worse than emulation, actually...

[No_not_like_Quake]

Commodore love isn't so bad. I've only played with Amiga emulators but was impressed.