[J-core] 2 CPU:s or nothing

BGB cr88192 at gmail.com
Fri Nov 25 23:16:36 EST 2016

On 11/25/2016 5:13 PM, Rob Landley wrote:
> On 11/25/2016 01:28 PM, BGB wrote:
>> On 11/25/2016 12:04 PM, Daniel V wrote:
>>> I'm not saying that this is what to do, I'm saying that it's a thought
>>> that may come in handy for building upon a possible future. It would
>>> draw people to that platform, instead of away from it, because it does
>>> not have any practical applications for them.
>> something with both 1 or 2 CPU cores and an MCU could be nice.
>> possibly, the MCU's address space and also IO peripherals are exposed as
>> a memory-mapped IO device, then the CPU can send a program over to the
>> MCU part and let it run, possibly with the MCU being entirely RAM,
>> rather than the usual SRAM+Flash (idea being here that the CPU loads a
>> program into it on startup).
>> maybe also they can...
>> possibly...
>> possibly...
> Feel free to prototype this. We released the code so people can do what
> they like with it.

could do it in the emulator easily enough, but I wont know the cost 
tradeoffs for a real FPGA.
still don't have a job or money, so sadly no money to buy an FPGA.

though, could at least detect if GCC steps on potentially omitted 
instructions (ex: if I omitted MAC and similar, would GCC notice?...).

>> otherwise, yeah, not been doing much here lately, have been off working
>> on other unrelated stuff, mostly on a newer Minecraft-style voxel
>> engine. not much directly relevant, but misc stuff:
> Good luck?

example video (some of what I have thus far):

started reviving it a few weeks ago, where initially I had wrote some of 
the core parts around a year ago but then went off and worked on other 
stuff (then did a platformer, but lost motivation, copy/pasted a lot of 
the code back into voxel engine after I revived it).

it was giving kind of crappy performance here, but I suspect this is 
mostly because (since upgrading to Win10) video capture from the desktop 
basically eats the CPU a lot worse than it did before, which also has a 
detrimental effect on the 3D engine. when not capturing it tends to get 
closer to a solid 60 fps (and still performs better overall than my 
older voxel engine).

I currently also have this under MIT terms, partly as I don't really 
expect anyone to care this time around either.

> I've spent weeks on GPS signal decoding, but it's not directly relevant
> here.


yeah, I work on various stuff, and often there is some overlap.

More information about the J-core mailing list