Getting closer...


This year, I doubt.

I couldn't say how much RAM it would have but if it's a package-on-package style SoC, it might have 4 or 8GB.

Most likely you will be able to buy a Pyra with the new board already installed, once it's available. But I wouldn't expect it to be this year. Maybe sometime next year or the year after. ED will need to make sure there is sufficient demand for it before he does anything.
 
We have seen how long it takes to make such an CPU board ready to run. So even if the development of a new one with a recent SoC would start today, it could took years until the finished board will work within the Pyra, sadly. Could be around the GPD WIN 5 release or so... ;)
 
Just use same SoC as RPi 4 (whatever that will be) for new Pyra SoC board ...
Which will probably be at best on par with the current Pyra CPU.

While it's hardly the best metric, Pyra Cortex-A15 is 3.5-4 DMIPS/MHz. Pi 3's Cortex-A53 is 2.3 DMIPS/MHz - about half the efficiency. That's the compromise you have to make for such a low price point. It is, however, 64-bit capable compared to the A15, which is only 32-bit capable, but right now this is a moot point due to the Pyra topping out at 4 GB of RAM. Any CPU with over 4 GB of RAM though would definitely need a 64-bit CPU of some kind, and obviously some apps could probably benefit from a 64-bit version, so logically speaking, it makes relatively little sense not to have an upgrade CPU be a 64-bit capable one.

If ED wants to keep to ARM processors, I'd think the ones he should be looking at would be Cortex-A72, Cortex-A73, or Cortex-A75. These all provide a notable increase in efficiency over the current Cortex-A15 (all being at least 5 DMIPS/MHz, with the A73 being about 6.35). The main questions would be how much he feels he'd benefit from stuff like per-core cache (A72 has the least), how much he feels he'd benefit from two-way vs. three-way pipelines (A73 has only two way, the other two have three), if he thinks he will go over four cores (only the A75 has support for 1-8, the others are strictly 1-4), and so on. Even then, all of this has to be balanced against stuff like heat produced, power consumed, and if it'd be a truly tangible boost to performance of the device - if it ups heat and battery consumption a lot but only gives about 20% more performance, is it worth it?

And even then, he's got to find someone who will sell them to him in bulk quantities. Qualcomm turned him down, nVidia isn't interested without full manufacturing control, and so on. It's not like desktops where you can just go to an online store and buy the processor you want.

Intel Atom would be the next-most possible chip, but of course, then we'd be taking a big leap up to x86-64. Debian does run on x64, so that won't be a problem, but it automatically doubles-up the amount of testing and maintenance that needs to be done, as well as that apps that would be compiled as DBPs would also need to specify what version they are (DBP = ARM, DBX = Intel), or be some kind of "universal" app (DBU) that would work on both. The other kicker is that at best, it'd turn the Pyra into a GPD Pocket clone. (Although some might not feel that's a bad thing.)

And no matter what, it's got to have a GPU - preferably with open-source drivers, because otherwise, we have to either hack and kludge around the GPU, or not have a device that's as open-source as he'd like it to be (which also prevents it from having more support than whatever its manufacturer decides to give it). Something like this could literally extend the life of the thing by years.

So in short, figuring out what the upgrade board should be isn't a cut-and-dry process. It's got to be something he can get in large quantities, if it's not ARM v7-A he has to support a whole other instruction set, there are binary concerns if he picks an architecture that's vastly different, what GPU it has and what can be done with it, and so on. It's not an easy choice to make.
 
Isn't Atom dead at the moment? All the tablet and portable makers seem to be using core-m3 chips for the time being.

Qualcomm might have failed to repond usefully to him at present, but once he already had a design in production they might be more forthcoming. Contacts will need to be reestablished, and we'll have to see what those present.

While it's fun for us to compare naked ARM cores, for ED and team it's only really worth thinking about SoC designs initially, then checking those against the single core performance of the ARM cores therein.

ARMV7-a versus ARMv8-a could provide interesting problems for the repo though, I think. IIRC the 32-bit mode of ARMV8-a is basically the same as ARMv7-a, but things designed to run in a 64-bit context will be incompatible. The dbp running tools will presumably need to know which context to switch into, but 32-bit ARMv8-a code could equially well run on an ARMv7-a chip.
 
@Yuji Sakai

Well ED responding or not, I'd say likely not.... Current CPU board has taken well over a year to iron out the issues. Currently the future SoC he's interested in isn't even released yet
 
For SOC I hope so much we will wait until there is an A75 available we can use.
According to Wikipedia:
The Cortex-A75 serves as the successor of the Cortex-A73, designed to improve performance over the A73 while maintaining the same efficiency.[2] According to ARM, the A75 is expected to offer 16–48% better performance than an A73 and is targeted beyond mobile workloads. The A75 also features an increased TDP envelope of 2 W, enabling increased performance.[3]
Even if it's only 20% faster than the A73 it's still a lot compared to the OMAP5.

And don't worry about the 2W. OMAP also uses up to 2W AFAIK.
 
For SOC I hope so much we will wait until there is an A75 available we can use.
According to Wikipedia:

Even if it's only 20% faster than the A73 it's still a lot compared to the OMAP5.

And don't worry about the 2W. OMAP also uses up to 2W AFAIK.

They mean up to 2W per core. I think an A55 based chip would have decent permance and efficiency. They would probably still be less performant per clock then the A15, but most of them will be clocked quite a bit higher.
Problem with the higher end cores is that they usually come in quite expensive SoC's. I'd still see an i.MX chip, if they would come out with an A55 based design as an ideal successor. But that would of course take time to come out and design.
I was thinking of waiting to buy a Pyra untill the newer SoC design was out, but I think I'll spring for one as soon as they come out.
 
While I don't see an upgrade CPU board next year, I do think the next iteration will be faster to design. We had some growing pains but now know how crucial simulation is to make sure our board stays in spec.

Building this way will save money and time in the long run.
 
The Novena laptop uses an i.MX as well. It's the best for open source but as was said earlier in the thread A53 cores are not exactly exciting. Depends on what the priorities are.

It isn't exciting if you look for performance, but it is much more exciting than OMAP 5 with Cortex-A15, because:

1/ It is a 64 bits CPU, instead of 32 bits.

2/ It isn't affected by Meltdown nor Spectre (Cortex-A53 is immune as it doesn't have speculative execution, while Cortex-A15/Cortex-A57 and higher ARM cores have it).

3/ It uses less power than OMAP5 and this is very important for a pocket computer.

If finally it has open drivers (in some areas), even better.

For me it is very exciting, even if in a mono-thread application it get lower score.
[doublepost=1517160867,1517159788][/doublepost]
A53 cores are not exactly exciting

Think twice. Cortex-A15 cores (2 of them) in OMAP 5 in actual Pyra is faster BUT, as they get hotter it will have to throttle down. Cortex-A57 cores in i.MX8 M will probably generate less heat for being simpler. So after a time running at full speed, specially in a hot day, we can end with OMAP5 being notably slower than it was initially.
 
Last edited:
I agree with most of your points and reasoning and I am not arguing that it is a bad chip.

Without actual figures on power usage and processing power (with all relevant security patches installed) it is difficult to have a definitive opinion anyway. We were just saying that this hypothetical SoC upgrade would probably not enable new emulated systems for example.

Edit: according to Wikipedia the i.MX 8M line is clocked at 1.5GHz. So if we take the ratio of 0.657 (according to the DMIP/MHz quoted for core A53 vs A15), we could throttle the A15 to 1GHz and still be at least on the same processing power level. I am not knowledgeable enough to know what other factors are important (caches, instruction sets, bus).

I wrote in my earlier message "Depends on what the priorities are.". For me game emulation is one and I'm guessing you may have over primary use cases for your Pyra.
 
Last edited:
I wrote in my earlier message "Depends on what the priorities are.". For me game emulation is one and I'm guessing you may have over primary use cases for your Pyra.

I understand you: if you use it to play emulators, for some of them, the more demanding, power is a must, because without that power you can't run them.

I love emulators and I use them, but I want this pocket computer as a universal pocket computer: I will use it to play, to work, to code, to surf the web on the go, etc, so I prefer 64 bits (I can run 64 bits soft), no Meltdown-Spectre bugs, better battery life, and more drivers in open format because in some cases if we don't have open drivers we can end with an outdated Linux kernel.

I don't want to sacrifice a more versatile and updated pocket computer to be able to run for example a Dreamcast emulator.

PD: I have been more user of computer games (8/16/32 bits) than console games, although I have console emulators of course.
 
This thread seems like the next CPU board is no huge upgrade, or did i skip too much?

If the next CPU board would be X86, the pyra would win against the GPD Win 2 in every category, at the same price.
With an ARM CPU, they are still not compareable.

So, using the same CPU than GPD is too much effort for the size of the Pyra project? GPD collected more than 1.500.000 USD, so there is actually a market.
 
This thread seems like the next CPU board is no huge upgrade, or did i skip too much?

If the next CPU board would be X86, the pyra would win against the GPD Win 2 in every category, at the same price.
With an ARM CPU, they are still not compareable.

So, using the same CPU than GPD is too much effort for the size of the Pyra project? GPD collected more than 1.500.000 USD, so there is actually a market.
Using a current Intel CPU would be a security liability. Also, it seems like ARM is the architecture of choice for the Pyra team.
 
I think it's important to do thermal simulations as well as physical tests to make sure we won't end like GPD did - where the CPU heat and battery charging heat are going over safety limits of the battery to the degree that it inflates. Recently there are numerous reports about it from those who use lots of high-performance tasks.

A few hints from my experience with simulations (FEM), but I work with much higher temperatures (metallurgy):
- Equivalent transfer coefficient from computations is usually assumed too high regardless of initial assumptions and computations made in preprocessing.
- If something is built with casing on all sides and is very thin, the best approximation is no external heat transfer.
- The materials heat capacity values in most software libraries are significantly exaggerated. I reduce them by computing coefficients from very simple porosity formulas, but this works for thermomechanical simulations.
- Timing predictions of any cyclic process, even based on the best measures, are usually terribly wrong, it's better to use the worst case of such simulations.
 
- If something is built with casing on all sides and is very thin, the best approximation is no external heat transfer.
Dependent on definitions of 'very thin', if that were true my Pandora along with a number of other electronic devices I own ought to be little puddles of white hot plastic by now. I suspect over short timescales this is a valid assumption, but it doesn't come close to holding as deltas increase and longer time periods are considered.
 
That looks like one to maybe evaluate, but judging by the mounting holes for a fan and the metal top to the SoC suggesting good thermal contact for a heatsink, that might run a bit hot. But maybe the Pyra's current cooling solution is better than I think it will be, and we can afford to use a hotter chip, we'll have to wait and see.
 

from here, it seems to run with lots of watts:

I measured the power consumption with or without SATA and/or USB hard drive(s) attached in power off, standby, and idle modes:
  • Power off – 0.0 Watt
  • Standby – 5.1 Watts
  • Idle – 5.1 Watts
  • Power off + USB HDD – 0.0 Watt
  • Standby + USB HDD – 8.1 Watts
  • Idle + USB HDD – 9.1 Watts
  • Power off + USB HDD – 0.0 Watt
  • Standby + USB HDD – 10.3 Watts
  • Idle + USB HDD – 11.2 Watts
 
Back
Top