Suggestion: better performing pinephone

[nodoze]

You're right that, if you want to run 3D games, the Pinephone is not a good fit for you.

This is an experimental device in an ecosystem (Linux applications on mobile devices) that barely exists yet. It uses a processor that is relatively tried-and-true for Pine64's development community - the A64 - which was a very reasonable choice because it keeps the system relatively simple to develop for and inexpensive. Computing power on the order of what other extant smartphones are capable of would have been a bad decision, because it would have jacked up the price and development complexity of a unit that is really not much more than a field testing platform for a very new category of free software, and not given us much in return because so little usable software is available presently.

If you really want a foss-phone that has more premium specs, you can always check out Purism's Librem 5. It costs 750 USD, and will be plagued with more or less all of the same software issues we have over here, just with overall slower progress, since far fewer people can actually afford to buy the phone in order to work on it. I like Pine64's approach better.

Now that the hardware in the Libram5 is working in Linux why couldn't there be a Pinephone version with the same CPU/GPU?

[marcih]

Now that the hardware in the Libram5 is working in Linux why couldn't there be a Pinephone version with the same CPU/GPU?

This is a genuine question with no ill will towards you: Why would there be? If you want the hardware of the Librem 5, you can buy it. I doubt PINE64 would be able (or willing) to manufacture the Librem 5 for cheaper or create a whole new design for the same hardware. Either way, I don't think the PinePhone's speed/responsiveness issues are in the (in)capabilities of the hardware as much as they are in the (im)maturity of the software.

Having bought the Manjaro edition, I can feel an improvement in performance and battery life in the last few months, now imagine what changes the people who have been here for all of last year have seen. I think we ought to Just Wait™ for a while longer before passing judgement on the PinePhone's hardware.

[ryo]

A Pinephone with a stronger CPU would nicely serve as a body warmer.
The current one is already getting hot pretty easily while screen is off and phone is in a pocket.

[AndrewBlackmiller]

Hi folks,

are you sure we need a "stronger" phone?
Maybe we just should invest more time and energy in software-optimisation to get maximum performance out of this little devil
I'm sure it's possible to get the software running 10 times faster just by optimization ... so would this fulfil your "need for speed"?

What do you think? Any ideas?

Let me start the list:

1) Lets optimize the UI-driver stuff for the Pine-Phone hardware.
2) The stage is yours .....

Best regards,
Andrew >:o)

[nodoze]

Hi folks,

are you sure we need a "stronger" phone?
Maybe we just should invest more time and energy in software-optimisation to get maximum performance out of this little devil
I'm sure it's possible to get the software running 10 times faster just by optimization ... so would this fulfil your "need for speed"?

What do you think? Any ideas?

Let me start the list:

1) Lets optimize the UI-driver stuff for the Pine-Phone hardware.
2) The stage is yours .....

Best regards,
Andrew >:o)

I think increased Software Optimization is great but still don't understand why having multiple options and/or choice is bad.

[AndrewBlackmiller]

Hi folks,

are you sure we need a "stronger" phone?
Maybe we just should invest more time and energy in software-optimisation to get maximum performance out of this little devil
I'm sure it's possible to get the software running 10 times faster just by optimization ... so would this fulfil your "need for speed"?

What do you think? Any ideas?

Let me start the list:

1) Lets optimize the UI-driver stuff for the Pine-Phone hardware.
2) The stage is yours .....

Best regards,
Andrew >:o)

I think increased Software Optimization is great but still don't understand why having multiple options and/or choice is bad.

On one hand because stronger usually means more expensive ...
and on the other hand I think we have to do something against these "planned obsolescence" problem.

If we use our resources - a mobile phone for instance - with more care we safe our environment in the long term and a lot of money in the short term -> win -win situation
For instance - would you buy a new car if just by doing a service it will run as fast as new?

Trust me, I know what's possible just by doing the right Software-Optimizations, because I've done this in the past many times on my dally job.
You wouldn't believe what's possible

And maybe you guys can add some further ideas to move forward in that direction?

Best regards from Vienna,
Andrew >:o)

[nodoze]

Hi folks,

are you sure we need a "stronger" phone?
Maybe we just should invest more time and energy in software-optimisation to get maximum performance out of this little devil
I'm sure it's possible to get the software running 10 times faster just by optimization ... so would this fulfil your "need for speed"?

What do you think? Any ideas?

Let me start the list:

1) Lets optimize the UI-driver stuff for the Pine-Phone hardware.
2) The stage is yours .....

Best regards,
Andrew >:o)

I think increased Software Optimization is great but still don't understand why having multiple options and/or choice is bad.

On one hand because stronger usually means more expensive ...
and on the other hand I think we have to do something against these "planned obsolescence" problem.

If we use our resources - a mobile phone for instance - with more care we safe our environment in the long term and a lot of money in the short term -> win -win situation
For instance - would you buy a new car if just by doing a service it will run as fast as new?

Trust me, I know what's possible just by doing the right Software-Optimizations, because I've done this in the past many times on my dally job.
You wouldn't believe what's possible

And maybe you guys can add some further ideas to move forward in that direction?

Best regards from Vienna,
Andrew >:o)

This is the 'Pinephone Hardware' subforum;
I will gladly use the 'Pinephone Software' subforum to make any Software Optimization recommendations I have there.

I won't be going into the 'Pinephone Software' subforum and telling threads that are pursuing software solutions &/or alternatives that hardware is the only acceptable question to be asked nor that Hardware is the only answer.

Some of us want to do things with Linux Phones that the Pinephone will never be able to do because of the hardware limitations no matter how well you optimize the software.  Some of us want true convergence and want to be able to do advanced containers, virtualization, & more which do require software but also require hardware that is expandable and it is not just about the Central Processing Unit (CPU) core speed...

My understanding is that at decently sized order quantities the i.MX-8M System on a Chip (SoC) chosen for the Libram 5 costs ~$20/chip while the A64 CPU chosen for the PinePhone is ~$5.

iMX-8M supports upto 8GB RAM (faster LPDDR4);
A64 only supports upto 3GB RAM (slower LPDDR3);

iMX-8M supports upto USB 3.0 (upto 10x faster);
A64 only supports upto USB2.0;

iMX-8M supports +20MP Cameras;
A64 only supports upto 5MP Cameras;

iMX-8M is a true 64bit SoC;
A64 is only 64bit on the CPU cores but 32bit elsewhere;

Graphics subsystems on the iMX-8M is several generations more capable than the A64...
iMX-8M supports 4K Video out at 60 FPS for convergence;
A64 only supports 1080p Video out;

etc...

A $15 increase in base CPU cost ($20-$5) would have allowed a much broader platform and spectrum of device options.   Sure start with a lowest cost option but not be stuck there...

I am all for low cost development options with minimal/slower RAM and cheaper initial components but wish the PinePhone wasn't locked into such low specs with no upgrade path and no options for increased specs.

[TRS-80]

iMX-8M

Choosing iMX-8 is why people who gave Purism $600 (or whatever) all the way back in 2017/2018 are just now beginning to receive their phones.  Let that sink in . jpg

I am all for low cost development options with minimal/slower RAM and cheaper initial components but wish the PinePhone wasn't locked into such low specs with no upgrade path and no options for increased specs.

I think "locked in" is a quite hyperbolic statement.  I also was happy to get 3/32 option (coming from Samsung s5 with 2/16) so I also do not know where you get "no options for increased specs"?  And the platform (Linux phones, I mean) is very, very young.  Who knows what may come in future?

I for one appreciate the strategy Pine have taken.  Linux phones have a long way to go before they are ready for Normies to use them as daily driver.  Therefore get as many of them into the hands of people to tinker / dev and improve the state of the platform overall.  I am actually thankful Pine have given us inexpensive alternative to Librem 5 because it means I can actually get one in my hands and start helping advance the state of Linux phones sooner rather than later.  And not just me but many other people, too.

[gamerminstrel]

I am by no means an expert, but the way I see it, the pinephone's hardware choices were made so they could be made available in such a larger scale, at an affordable price when anything better would just be $$$$$ and would run too hot or used closed-source blobs.



By the time better hardware becomes available, enough software advancements will have been made so that this current pinephone will run like a champ. My hunch is that the pinephone 2 in maybe 3-5 years will be able to take advantage of all the progress right out of the gate on top of beefier hardware specs, so it will feel like a true generational jump in improvement.



Just give it time, but once maybe Manjaro or Mobian have stabilized/improved enough, I do think even some 3D indy games could be officially ported to (or even made for) pinephone.

[dsimic]

Quite frankly, we should all be very thankful for the fact that PinePhone exists at all, especially at such a low price.  We should actually be happy that it uses old Allwinner A64 SoC, which has been around a while, and as such is pretty well known inside out.  That allows good software support (i.e., mainline Linux kernel support), which is still work in progress.

However, I do agree that a die shrink of A64 would be great, with lower power consumption and all, but that's probably never going to happen because it would have no economic sense to Allwinner.  Unless Pine64 manages to ship millions of PinePhones and pursuades Allwinner to invest into a die shrink.

The way I see PinePhone is a nice, open embedded computer that can also be used as a phone.  I would never expect PinePhone to run any 3D games, simply because that's not reasonable.  The market for handheld game consoles (or highly capable smartphones) is already overcrowded, so selecting one to enjoy some games should be no problem.

Furthermore, having not so powerful hardware as the target platform is actually beneficial for the software quality.  Poorly written software runs fast on fast hardware, but only very well written software runs fast on slow hardware.