ARM vs. X86. Lets vote

[God Ginrai]

Yes, but the majority (not all, but the majority) of optimized ARM applications have also had important UI tweaks that will better fit them on a mobile device. On x86, this would not be the case, since x86 has not had much hold in the mobile space until now.

-God Ginrai

Again - why would any of this not be portable to an x86-based platform? Most Android stuff already runs on x86. Most Pandora stuff is open source and portable to x86 (although that doesn't stop a lot of them from using desktop-like interfaces). I'm really not seeing how this is a problem tied to architecture.

Actually, I was talking about from the other end of the stick. These applications that you were talking about for x86 machines that were less powerful than Baytrail would need UI optimizations for mobile use. That's what I was talking about would waste time. And if some of those are closed source, obviously you don't even have that option. (like the Steam games)

So - you think that any ARM CPU, regardless of technology, availability, driver support, performance per watt is favorable regardless of competition?

No room for rational reason, facts, options or alternatives.

Oh come on, you know he didn't mean that. Stop trying to twist his words.

-God Ginrai

 

[TrashyMG]

So - you think that any ARM CPU, regardless of technology, availability, driver support, performance per watt is favorable regardless of competition? No room for rational reason, facts, options or alternatives.

No I'm just stating what I'd like to see as an outcome, I've made view clear on a prior page.
http://boards.openpandora.org/index.php/topic/14399-arm-vs-x86-lets-vote/page-4#entry276657

 

[Exophase]

Most Android stuff already runs on x86

 [citation needed]

People seem to disagree on this one....

What proportion of popular Android software actually uses ARM-specific code (and doesn't have an x86 version hiding around the corner)?

According to Intel, 90% of ARM-only code works with their binary translation in Android. I don't believe they fudged that number, although "works" doesn't exactly mean "works well." I've read reviews of several x86 Android devices and reports of some app not working are not common - although only a few of them bothered to mention that they ran dozens of apps to test this. For what it's worth, DraStic explicitly doesn't support x86 right now ;p

So you say people disagree, who is actually reporting something to the contrary? And not just making a baseless claim.

 

[Binky]

So you say people disagree, who is actually reporting something to the contrary? And not just making a baseless claim.

None that I remember.

Its just that (claim vs claim) is harder to draw a conclusion from that (claim vs citation)

Now someone has actually provided specific data, its possible to draw a conclusion.

 

[fusion_power]

Such a vote actualy needs more infos. What ARM chips are possible, what X86 chips are possible? An more detailed facts sheet would be also a good idea to compare the ARM and x86 leagues. pro's/con's etc... Otherwise it would be just an question of your personal preference instead of the choosing what's the BEST for Pandora 2.

 That's what I intended.

Then I have to vote "Don't care", I have no special preference only want to be sure it's the best choice for the Handheld. Maybe it will be ARM, maybe the new "mobile" x86" wonderthing that was mentioned, who knows. Whatever ED will choose, I'm sure he will check and double check that the new Chipset will be a good one, for devs, for battery life and for availability.

 

[Linux-SWAT]

We can always found defaults, but anyway ARM > x86.

I don't know what's with you, you cry foul when a thread doesn't meet your quality standards but continue to make post after post like this >_> This doesn't help the level of common knowledge about this topic here, which is currently full of misconceptions.

And btw, x86_64 is better, because it was made by AMD.

Your Intel hate is clearly clouding your judgement. x86-64 was a pretty pedestrian and by the books change. x86-64 is better because it's newer (AVX and AVX2 make a lot of fundamental improvements too, and those aren't AMD). It wouldn't have made sense in 1985 when Intel made their last big mode change. Intel had little real problem going along with it. AMD didn't bring any unique magic to the table.

A cow produces more milk than a horse, but a horse can jump higher. Of course usage may lead to choose a cow over a horse. In our case, ARM is better suited.

I used c64, mo5 and to7-70 long before PC so when it comes to look at x86 as one amongst many other, i must say that if AMD never came with their K7 architecture, we would currently still be using P3 or P4. There was a time Intel and AMD machines were in equal numbers in shops (t-bird and after). Then AMD pushed 64bits, blasting the expensive Intel's IA64 and forcing them to use their money, let's say, to reduce nm manufacturing with a lot of $$$ owned with the wintel racket. Choice matters, not only products advantages. Not supporting Intel has proven good.

 

[Exophase]

A cow produces more milk than a horse, but a horse can jump higher. Of course usage may lead to choose a cow over a horse. In our case, ARM is better suited.

I used c64, mo5 and to7-70 long before PC so when it comes to look at x86 as one amongst many other, i must say that if AMD never came with their K7 architecture, we would currently still be using P3 or P4. There was a time Intel and AMD machines were in equal numbers in shops (t-bird and after). Then AMD pushed 64bits, blasting the expensive Intel's IA64 and forcing them to use their money, let's say, to reduce nm manufacturing with a lot of $$$ owned with the wintel racket. Choice matters, not only products advantages. Not supporting Intel has proven good.

I'm still waiting for anything resembling a technical argument from you. I probably shouldn't be holding my breath..

The idea that we'd still be using Pentium 3s or 4s if not for AMD is cute, but thoroughly absurd. Intel continually developed new processors before AMD was competitive and continually developed new ones after they were competitive. If they stopped doing this people would no longer replace their Intel processors with other Intel processors, at least not until their computers broke down. It's funny how you seem to even suggest that AMD inspired Intel to develop superior manufacturing. Moore's Law has been a near constant for decades.

I won't deny that Intel has abused the market and that they've been tied to more restrictive platforms (not that AMD is better in that regard). But none of that has anything to do with whether or not their current mobile processors are suitable for a Linux handheld. But I'll bet you're one of those people who still insists on using AMD because you have a grudge against Intel and not because they have better processors... This is all right up there with your cries that no one uses Windows anymore.

 

[Linux-SWAT]

Even if i don't know low level CPU stuff, overall knowledge about history may suffice. Intel was choosen by IBM because they were the cheapest (as m$, which was simply ms at this time). For the record, not only that: they even ordered a lower model than 8086 for the PC. The child was crippled from the beginning. And no matter the efforts, you can't rewrite, well, genetics.

When you have the monopoly, you don't need to push hard. For example, No AMD, no cheap x86_64. It's not absurd, it's how greedy companies work, just look around.

Intel is better than AMD on high budget, which i don't have. But if someone ask me for the fastest no matter the price, i'll say Intel. I have no problem with that.

On another side, based on my own experience, i found AMD virtualization with KVM far better. Milk and jump.

Look around (again), you'll see far more ARM/UNIX phones/tablets than wintel.

Yeah, there's still big companies enslaved to Office Pro Premium Full Light Family Edition.

You're tied to your engineer pov, i'll tell ya.

 

[Exophase]

This stuff that happened 30 years ago with Intel does not matter. You're wrong about no AMD meaning no cheap x86-64 - because Intel stopped worrying an awful lot about competing with AMD a couple generations ago. If anything, AMD constantly teetering on the edge of bankruptcy but not yet falling over is doing Intel a favor so they don't need to worry about attaining a monopoly status. Like I already said, the market dictates price based on what people are willing to spend to replace the parts in their computers, and this has gotten more stringent as people become less interested in upgrading (despite inflation high-end CPU prices have largely gone down). It doesn't matter how greedy they are, they price parts in a way that'll sell.

Everyone knows that Windows is barely used in phones and tablets. It doesn't mean that people aren't using Windows on their desktops and laptops, and yes people are still using desktop and laptops. Not just big companies enslaved to Office. It's sad to see people using that "Wintel" term to mean that you can't have one without the other. That duopoly is done. You'd think someone in support of choice would recognize that.

The bottom line is, none of the stuff you're saying is making a compelling argument on why a Pandora succesor should avoid BayTrail. If that's my engineer-POV talking I'm fine with that.

 

[Linux-SWAT]

AMD was in a very good position during the t-bird/athlon-xp era.

People are still very interested in upgrading. Their phones. iPhone 5, how much does it costs ?

Maybe a real-life BayTrail benchmark may change my mind, but it's probably not gonna happen.

 

[thatgui]

Microsoft introduced its new Surface 2 Tablet, I wonder why they chose the Tegra 4 over a Intel "somethingtrail" ?* Availability, pricing, is WinRT only ?

*Disclaimer: this is not meant as a ARM vs. Intel argument, I'm just curious

 

[Exophase]

Microsoft introduced its new Surface 2 Tablet, I wonder why they chose the Tegra 4 over a Intel "somethingtrail" ?* Availability, pricing, is WinRT only ?

*Disclaimer: this is not meant as a ARM vs. Intel argument, I'm just curious

My guess - WinRT wasn't designed for x86 processors, and putting a full Windows tablet at Surface 2's price would confuse the market segmentation they're establishing with the much more expensive Surface Pro 2. Surface RT was a flop in no small part because no one's really interested in Windows if it means not getting legacy compatibility or the desktop. Surface 2 doesn't change that, so it'll be interesting to see how MS expects it to do better. But it does seem like the pricing is meant to deliberately inflate the value of Surface Pro 2. Not really great for a tablet that's not especially cheap.

I could you could argue that Windows RT has built up some app market that needs ARM but probably not much of one...

 

[vcoleiro1]

Microsoft introduced its new Surface 2 Tablet, I wonder why they chose the Tegra 4 over a Intel "somethingtrail" ?* Availability, pricing, is WinRT only ?

*Disclaimer: this is not meant as a ARM vs. Intel argument, I'm just curious

Windows RT runs on ARM SOCs , which is why the Surface 2 uses a Tegra 4 and not Bay Trail.

Surface 2 Pro runs full Windows 8.1 Pro and has a x86 Core i5 Haswell CPU in it.

MS seems to have gone for premium options in both ARM and X86 (at least in SOC/CPU choice)

Other manufacturers have all but deserted RTand are releasing budget Windows 8.1 tablets using Bay Trail.

ie Acer, Asus, Dell, Fujitsu, HP, Lenovo, LG and Toshiba are all developing Bay Trail based Windows 8.1 tablets

The first Bay Trail - Win 8.1 tablets are due October / November with screens ranging from 7 - 11.6 inches . Some/All will include Office

So it wont be long until you see what they are like in real life.

 

[_wb_]

Given those perf/W numbers, and assuming that it will have good FOSS support, I now say: go BayTrail!

I don't like Intel as a company, but then again I don't like Apple or Samsung etc either - all big companies tend to be evil. At least Intel seems to be embracing FOSS somewhat in recent years.

In terms of technical niceness of the instructions sets: yes ARM is nicer than x86, but I personally don't write assembler code so I don't care, as long as gcc does a decent job at writing the assembly for me

Breaking backwards compatibility is not so nice, but it's not the end of the world. Also, if internal storage is big enough, we should be able to install most of the software as regular packages from whatever distro we use, and just use that packaging system instead of PND for a lot of the software (e.g. web browsers, LibreOffice, etc). We can of course still use the PND system where it makes sense (e.g. for games), and either make a new repo for the P2, or adapt the PND spec to support multiple architectures.

 

[Askarus]

Most Android stuff already runs on x86

 [citation needed]


People seem to disagree on this one....


What proportion of popular Android software actually uses ARM-specific code (and doesn't have an x86 version hiding around the corner)?

According to Intel, 90% of ARM-only code works with their binary translation in Android. I don't believe they fudged that number, although "works" doesn't exactly mean "works well." I've read reviews of several x86 Android devices and reports of some app not working are not common - although only a few of them bothered to mention that they ran dozens of apps to test this. For what it's worth, DraStic explicitly doesn't support x86 right now ;p


So you say people disagree, who is actually reporting something to the contrary? And not just making a baseless claim.

I don't know why X86 compatibility is so important for you just because of closed source stuff.

For me 100% Android support would be the major point.

In my opinion Android as an OS is crap but also a mobile device that can't run Android apps is bad marketing.

"Hey, I have a awesome fast Mini PC but it can't run what your Smartphone can"

With ARM Linux we have nearly all the programs we have for desktop Linux (at least the ones I want) AND Android.

With x86 we have (maybe) a much bigger Linux repo but no Android stuff.

And you always have to keep in mind that we don't have to use a tiny Angstrom Linux.

We may use a Debian or Arch Linux and there's lots of packages in their ARM library already.

(See PanDebian)

 

[Linux-SWAT]

A fresh (such a coincidence) french article about the 10 years of the Athlon 64:

http://www.hardware.fr/news/13321/athlon-64-10-ans-deja.html

 

[rohezal]

With x86 we have (maybe) a much bigger Linux repo but no Android stuff.

Exo said you would have 90% of the Android software.

But we would need something like autobuild in debian. That every open source software is uploaded as source code, automaticly builded by the repo and packaged into a PND. This way there would be always a Pandora and Pandora 2 version (except people don't want it for both devices for reasons like NEON code or  the use of full open gl 3).

The instant avaibility of the old stuff is big pro for ARM, the use of x86 linux and windows games with a gamepad (the Pandora) and wine is a big pro for X86. Imagegine Thief, Gothic, Oblivion etc to got. This would be a great thing and a big seller.

 

[ekianjo]

With x86 we have (maybe) a much bigger Linux repo but no Android stuff.

WHat are you talking about ?? Android can run on x86 as well ! It's common knowledge nowadays, and you can definitely use it on a x86 architecture. There's not much ARM specific stuff there, most of Android runs on Java layers.

 

[WizardStan]

With x86 we have (maybe) a much bigger Linux repo but no Android stuff.

Exo said you would have 90% of the Android software.

Exophase said that Intel said that 90% of ARM specific Android code would work. There are so many ways to interpret that statement and considering its source I wouldn't automatically assume it is the most favourable one.

 

[Linux-SWAT]

So again, real-life benchmarks are still missing.