The Future of Pyra's CPU


Grench said:
By that time anything 'new' would so vastly out perform the existing SoC that it would be akin to re-purposing an 8086 desktop machine (computationally not worth the cost of electricity to keep it powered on).
Click to expand...

That is both realistic and depressing at the same time.

Buuuut.... If I realy want to?!
 
I like this idea. Already when I get a Pyra, my old Pandora will be re-purposed for TV-out or 2-player console emulation (over LAN along with Pyra). If and when a newer CPU board upgrade is released, it would be awesome if the old CPU board could be used to upgrade a TV into a really smart TV (with a full Linux OS on it) and at that point, replace the Pandora if mine is still functional by then.

At which point the Pandora becomes a hand-me-down to someone who lacks a handheld game system but would very much appreciate one.

Color me interested.
 
I thought Ed said he was gonna release source code so anyone could make new fast cpu or improvements once the pyra was officially released so it wouldn't take so many years to get one


Sent from my iPod touch using Tapatalk
 
2coop4 said:
I thought Ed said he was gonna release source code so anyone could make new fast cpu or improvements once the pyra was officially released so it wouldn't take so many years to get one


Sent from my iPod touch using Tapatalk
Click to expand...

He already released the electrical schematics, but not sure who would tackle making a new CPU board outside of anyone in this community, it's no easy task for sure...
 
Oh I must of missed that post then. There is a lot happening in my life at the moment can't see everything at once


Sent from my iPod touch using Tapatalk
 
TrashyMG said:
@hns has already stated he wanted to build a board that could be used with the Pyra CPU modules.
Click to expand...
I think he had a tablet on his site I was thinking would benefit from that.

I would personally want a tablet (standard or e-ink display) that resembles a PADD...probably a red one. ;)

A functional tricorder would be cool, and it could include a camera and be a Pyra accessory. Call it the Pycorder or Pyracorder, I guess, but I know this isn't really a low development cost device like a tablet. Would be fun, though.

As long as it is kinda cheap, I would probably get something to use with a tv and the unused SoC.
 
I'd like to see an 64 bit arm chip at some point. Core count doesn't matter a whole lot to me, but it would be cool to get a more power board that uses less power :) /dreams
 
Grench said:
In the hypothetical future, 4 years or so from Pyra release, if a new SoC is released; on that day the existing Pyra SoC becomes nearly worthless and may as well be used as a keychain decoration. By that time anything 'new' would so vastly out perform the existing SoC that it would be akin to re-purposing an 8086 desktop machine (computationally not worth the cost of electricity to keep it powered on).
Click to expand...
Actually, one of the reason why I'm paying attention on this project is because of the CPU it use. The OMAP 5 while being interesting as it is, is also a rather aged product that half decades have been passed since its release. While its actual performance is unclear due to lack of usage on consumer electronic, other cpu with similar architecture are already hard pressed to handle some of the latest mobile app and games. Should also be less powerful than the portable computer released by the certain Chinese company a while ago. It would be nice if the cpu can be upgraded from it now but OMAP 5430/5432 is already last product in the line released by TI and any upgrading would mean switching manufacturer and redoing everything related to it and the extra time and effort and cost....
 
qunow said:
It would be nice if the cpu can be upgraded from it now but OMAP 5430/5432 is already last product in the line released by TI and any upgrading would mean switching manufacturer and redoing everything related to it and the extra time and effort and cost....
Click to expand...
Actually, that part shouldn't be as big of an issue as you might be thinking. By making the SoC board modular - and that it includes the RAM on the SoC module, THAT part would need to be redesigned to match up to a future SoC. The 'main board' that has the keyboard on one side and the SDXC slots on the other is relatively 'dumb'. It's systems, for the most part, trunk to the connectors for the SoC module. In theory, any future SoC that can handle at least the I/O lines to the 'dumb' mainboard AND fit in the thermal profile of the device could work.

Designing a new SoC/RAM/eMMC daughter board is NOT insignificant. But - it is one component that should be able to be redesigned while keeping the rest of the hardware 'as is'.

I could see this evolving into a 'leap frog' effect. The Pyra comes out, gets wide enough adoption to justify it, gets a new SoC board. It will only be able to take advantage of the features already present in the main board. The main board & case can then get an 'upgrade' to fully take advantage of the new SoC's feature set (which may include things that haven't been invented yet). The combination of the two would be a Pyra II? But the point is that by having the system modular, it is hypothetically possible for each piece to receive upgrades independently.

Where I expressed skepticism above was to the idea of re-using the 'old' Pyra SoC componnent 3-5 years from now in a non-Pyra enclosure after an upgrade SoC has taken it's place in the Pyra. By then, a functioning used Pyra would likely have less value than a dumbed down box would - which would negate the need for a dumbed down box (with it's development costs) to house the module.

That doesn't preclude the Pyra from having 'sister devices' that use the same modular SoC setup - those could make sense far before the Pyra gets an SoC upgrade part.
 
Yes, the next CPU board will probably be from a different manufacturer, but don't underestimate the amount of work done on the parts that won't change - selecting an LCD, touchscreen that doesn't cloud the display, keymat, d-pad, designing the way to squeeze in four shoulder buttons, six main game buttons, the main board. It's only the CPU board and maybe a few drivers in the OS that will need to change - even if we go 64bit, those CPUs still ship with a 32 bit mode which should enable widespread compatibility.

Edit: Ninjaed by @Grench, with a far longer answer.
 
Grench said:
Actually, that part shouldn't be as big of an issue as you might be thinking. By making the SoC board modular - and that it includes the RAM on the SoC module, THAT part would need to be redesigned to match up to a future SoC. The 'main board' that has the keyboard on one side and the SDXC slots on the other is relatively 'dumb'. It's systems, for the most part, trunk to the connectors for the SoC module. In theory, any future SoC that can handle at least the I/O lines to the 'dumb' mainboard AND fit in the thermal profile of the device could work.

Designing a new SoC/RAM/eMMC daughter board is NOT insignificant. But - it is one component that should be able to be redesigned while keeping the rest of the hardware 'as is'.

I could see this evolving into a 'leap frog' effect. The Pyra comes out, gets wide enough adoption to justify it, gets a new SoC board. It will only be able to take advantage of the features already present in the main board. The main board & case can then get an 'upgrade' to fully take advantage of the new SoC's feature set (which may include things that haven't been invented yet). The combination of the two would be a Pyra II? But the point is that by having the system modular, it is hypothetically possible for each piece to receive upgrades independently.

Where I expressed skepticism above was to the idea of re-using the 'old' Pyra SoC componnent 3-5 years from now in a non-Pyra enclosure after an upgrade SoC has taken it's place in the Pyra. By then, a functioning used Pyra would likely have less value than a dumbed down box would - which would negate the need for a dumbed down box (with it's development costs) to house the module.

That doesn't preclude the Pyra from having 'sister devices' that use the same modular SoC setup - those could make sense far before the Pyra gets an SoC upgrade part.
Click to expand...
I see, thanks for the explanation!
 
So I was thinking about this modular approach and component upgrade concept.

In theory, the only piece of the Pyra that shouldn't need to be 'updated' at any point would be the keymat. Replaced if/when worn out - yes. But...

SoC can be upgraded.
Main board can be upgraded - conceptually with new/different ports requiring new lower case plastics.
Keyboard surround/hinge can be upgraded.
Screen can be upgraded.

In theory, though, the keymat and keyboard layout could live forever through any other hardware upgrade cycle. Good thing we spent so much time on it.
 
Grench said:
Main board can be upgraded - conceptually with new/different ports requiring new lower case plastics.
Click to expand...
We should have put one of those port plates on the back:
blanking-plate-against-motherboard-ports.jpg

A missed opportunity :p
 
The needs of the modern world to solve problems using deep learning and real-time low latency processing are increasing compute workloads in the enterprise, and migrating tasks into the edge and end point.
Click to expand...
There are not the SoCs you are looking for; wavy hands.
 
You must log in or register to reply here.

Similar threads

ToastBucket
Future of the CPU board?
9 10 11
Replies
218
Views
56K
M-HT
M-HT
1
Pyra's new cpu board soc.
Replies
4
Views
3K
levi
levi
L
Using Pyra CPU Board for other purposes
Replies
12
Views
3K
Ligachamp
L
pmprog
FPGA CPU Board
2
Replies
20
Views
9K
Linux-SWAT
Linux-SWAT
Pyramancer
Location of the Pyra's mobile functionality?
Replies
11
Views
4K
Pocak
P
Back
Top