Discussion in 'Pyra News' started by EvilDragon, Jan 23, 2018.
Yep, if they're fine, they'll produce the 500 clear cases and we'll move the molds.
i'm want a powerpc one (XD) or, more realistic and even better, a RISC-V... in 3 or 4 years after the pyra is shipping
Yep, let's see how RISC-V behaves in real life and if it gets a (decent) graphics chip, but this would definitely go the right way regarding FOSS and OHW
Let's start our own OHW GPU design. Maybe it'll be done in time for Pyra Upgrade MarkV.
All we need to do is figure out how to shrink one of these.
How about a FPGA CPU board?
Do they have open programming interfaces?
Yep, some of them. https://www.olimex.com/Products/FPGA/iCE40/
i'll only buy it if it's a 64-byte instruction set.
How much will they cost? - I need a housing for my Pre-Preorder PCB!
Brings back memory. My first 3d card was a Voodoo type and i loved it!
Edit: Actually not entirely true. First was Matrox Mystique but that had crappy 3d so i slapped Voodoo 2 addon card. and hardware DVD decoder card later on. Bid daisy chain lol.
Depending on how much they cost I might want to get a clear case!
Watch clear cases sell out before I order one or two.
To me the Pyra would be versatile enough to motivate a low power SoC to keep it running longer at some reasonable cost of performance. My use case is more productivity focused rather than gaming focused but I would of course still want to play games on it.
So for me I'd jump on an i.MX8-M board to get the hopefully hassle free etnaviv driver with mainline kernel support. With moderate clocks and half of the cores isolated I would still be able to code and look up papers/manuals with a decent performance margin. When I want to play games I can easily reclock and taskset the game for maximum performance.
All in all I think that even as a sidegrade performance wise it makes sense to fill the different use cases. It's sort of the beauty of open software and hardware that you can optimize it to suit exactly what you want instead of being a good compromise for the majority.
If it makes sense economically is a different question and I suspect that's where the difficult issues are; what sort of returns on investment, if any, would you get from developing alternate boards and selling them to such a small audience? Will the SoC manufacturer even bother with such low quantities?
The Pyra is aiming for similar battery life to the Pandora with the current SoC. The Pandora had a great battery life. A new SoC should focus on more powerful processors, which should hopefully allow us to do the same for less while still opening up the possibility of broader horizons in software.
For an upgrade, I agree. For a sidegrade I think trying to improve battery life is a great option. Not everyone needs pure power. Once we have had them in our hands for a while and can see where each of us feels we want improvements I think it would be nice to have a couple, possibly unofficial, new CPU boards. Making it through 16 hours with 4G on, a few hours of real use and frequent checks with the screen on, would be awesome.
If we can improve both, or at least not really lose battery life while still improving processor power, I am totally for that.
Honestly, I don't see any reason you need more than 10 when you have a removable battery. (and when most everyone has a power bank nowadays) Plus, the Pandora had ~10 hours of real use. Why do you think the Pyra would be worse? I would think the Pyra should be able to hit a few hours of real use and frequent checks for a total of 16 hours without a different CPU board.
It's not going to be cheap for ED to develop a new CPU board; He needs to aim for something that the majority of Pyra users will be interested in upgrading to, or it won't be worth development costs.
I don't myself see how a removable battery helps if you have poor battery life. Shutting down even a phone so you can swap the battery is a pain, and if I know I'm going to be without power for a good while I'll just carry enough power packs to recharge it, even though that is a more inefficient process. Removable batteries are essential when it comes to keeping a device working beyond its design life, and my phone is on its third battery now for that reason, although assuming the Pyra battery is as good as the one still in my Pandora, and still rocking on, perhaps having screwdriver-free access to it is actually a luxury for my use cases.
How the hell is 10 hours of real use a poor battery life?
Separate names with a comma.