[ElReg] ARM tests: Intel flops on Android compatibility, Windows power

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
Not an entirely neutral article is it? Its a pure copy/paste from ARM marketing.
 

NTMBK

Lifer
Nov 14, 2011
10,377
5,521
136
Not an entirely neutral article is it? Its a pure copy/paste from ARM marketing.

Oh, certainly take it with a pinch of salt. The Windows testing certainly seemed a bit suspect; I'm not sure how much commonality there is between Windows RT and 8.1, and the lack of performance numbers was pretty telling. But the stuff about the binary translator is what I was more interested in to be honest.
 

mrmt

Diamond Member
Aug 18, 2012
3,974
0
76
I'd say that the relevant fact here is ARM worrying enough about Intel to make these slides public, since whatever is there is already known by their OEM partners in much deeper details.
 

NTMBK

Lifer
Nov 14, 2011
10,377
5,521
136
I'd say that the relevant fact here is ARM worrying enough about Intel to make these slides public, since whatever is there is already known by their OEM partners in much deeper details.

There is also an element of customer "education" involved here, don't forget. If ARM can successfully spread the word that some of your apps won't work properly on an Intel phone/tablet, it could have an impact on consumer sales. Nothing like a bit of FUD to keep your customers in their place ;)
 

Arkadrel

Diamond Member
Oct 19, 2010
3,681
2
0
There is also an element of customer "education" involved here, don't forget. If ARM can successfully spread the word that some of your apps won't work properly on an Intel phone/tablet, it could have an impact on consumer sales. Nothing like a bit of FUD to keep your customers in their place ;)


You got caught out by bad choice of words there.

"some of your apps wont work properly on an intel phone/tablet"

Isnt that the truth though? are you saying intel has a 100% success rate when it comes to emulating ARM? on x86.

Even if its only 6-9% as long as that magin of error is there, "some" is a fitting word.

I wouldnt call that FUD.
 

Haserath

Senior member
Sep 12, 2010
793
1
81
It all comes down to how good Intel's 14nm chips are. If they're so good you can't pass them up, they have a shot.

The 9% of apps that don't work are probably not what you'd use anyway... "Feed the baby" or something.:hmm:
 

NTMBK

Lifer
Nov 14, 2011
10,377
5,521
136
You got caught out by bad choice of words there.

"some of your apps wont work properly on an intel phone/tablet"

Isnt that the truth though? are you saying intel has a 100% success rate when it comes to emulating ARM? on x86.

Even if its only 6-9% as long as that magin of error is there, "some" is a fitting word.

I wouldnt call that FUD.

Oh, I'm using FUD in the strict sense of Fear, Uncertainty and Doubt- which is certainly what would be in my mind if I was considering an x86 Android device. It's certainly true.
 

NTMBK

Lifer
Nov 14, 2011
10,377
5,521
136
It all comes down to how good Intel's 14nm chips are. If they're so good you can't pass them up, they have a shot.

The 9% of apps that don't work are probably not what you'd use anyway... "Feed the baby" or something.:hmm:

This was taken from the top 25 free and paid apps, not barrel scraping trash.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
http://www.theregister.co.uk/2014/05/02/arm_test_results_attack_intel/

Looks like the binary translator still needs some work... :( Kind of disappointing to see the number of incompatible apps growing, not shrinking.

Relevant or not, history unfortunately shows that binary translators are not the panacea of opening a market to an otherwise outsider.

I wish Intel the best, but they have to do something "more than..." if they want to crack this nut.

ARM knows and is counting on this, FUD is their first and best card to play at this time.

But one has to wonder, with Intel's top-notch crack design teams, wouldn't a custom designed ARM chip (ala AMD's goal) developed by Intel's design team for Intel's world-class process nodes be the bomb in the world of anything ARM?

Intel is trying their best to make a square peg fit into a round hole, but they'd do so much better developing a world-class round peg to fit into that round hole. So why not?
 

scannall

Golden Member
Jan 1, 2012
1,960
1,678
136
Relevant or not, history unfortunately shows that binary translators are not the panacea of opening a market to an otherwise outsider.

I wish Intel the best, but they have to do something "more than..." if they want to crack this nut.

ARM knows and is counting on this, FUD is their first and best card to play at this time.

But one has to wonder, with Intel's top-notch crack design teams, wouldn't a custom designed ARM chip (ala AMD's goal) developed by Intel's design team for Intel's world-class process nodes be the bomb in the world of anything ARM?

Intel is trying their best to make a square peg fit into a round hole, but they'd do so much better developing a world-class round peg to fit into that round hole. So why not?

I'm sure Intel could build an amazing ARM SOC. I'm guessing the reason they don't is that ARM is a competing ISA, and one that is open. Open in as much as anyone can license from ARM, they'll take anyone's money. Not a good thing for Intel.

As long as ARM is strong in mobile, Intel can't get their usual margins for their products in that space.
 

Shivansps

Diamond Member
Sep 11, 2013
3,897
1,557
136
binary translators are a kind of a pain, my first tablet was a MIPS JZ4770, i had to use a program that translated ARM libs to MIPS, and i also had to take matters intro my own hands to recompile a few programs myselft.
Still, a lot of programs never worked.

From what ive heard, Intel translator is a lot better, but chances are, you gona still run intro incompatible apps, and the number will be increasing fast because apps are becoming more complex.
 
Last edited:

SammichPG

Member
Aug 16, 2012
171
13
81
There is also an element of customer "education" involved here, don't forget. If ARM can successfully spread the word that some of your apps won't work properly on an Intel phone/tablet, it could have an impact on consumer sales. Nothing like a bit of FUD to keep your customers in their place ;)

It's not fud when some apps aren't compatible and you incur in performance/battery hits because running a binary translator isn't free.

ARM is the incumbent in this market.


Relevant or not, history unfortunately shows that binary translators are not the panacea of opening a market to an otherwise outsider.

I wish Intel the best, but they have to do something "more than..." if they want to crack this nut.

ARM knows and is counting on this, FUD is their first and best card to play at this time.

But one has to wonder, with Intel's top-notch crack design teams, wouldn't a custom designed ARM chip (ala AMD's goal) developed by Intel's design team for Intel's world-class process nodes be the bomb in the world of anything ARM?

Intel is trying their best to make a square peg fit into a round hole, but they'd do so much better developing a world-class round peg to fit into that round hole. So why not?

Corporate culture?
Embrace/extend/extinguish by using x86 as a barrier of entry once they gained market?
 
Last edited:

NTMBK

Lifer
Nov 14, 2011
10,377
5,521
136
It's not fud when some apps aren't compatible and you incur in performance/battery hits because running a binary translator isn't free.

ARM is the incumbent in this market.

Again, I was in no way saying that it was incorrect :) It's pretty clearly (and demonstrably) true. I was going off the original Amdahl definition:

FUD is the fear, uncertainty, and doubt that IBM sales people instill in the minds of potential customers who might be considering Amdahl products.

The information does not in any way have to untrue- I am only classifying it a FUD as it sows the seeds of Fear, Uncertainty and Doubt into the minds of anyone contemplating a switch from ARM to Intel. Pointing out real world failings is a very good way of doing that- they Fear incompatible applications, are Uncertain about the performance of the applications that do actually run on the binary translator, and they Doubt whether Intel's performance claims will actually translate into real world improvements.
 

Madpacket

Platinum Member
Nov 15, 2005
2,068
326
126
Intel's definitely not going to break into the Android market any time soon by trying to shoehorn X86 on it, even at 14nm. Binary translation is simply not an elegant solution and developers really don't need to care about supporting X86 in the tablet / smartphone market to make money.

Intel's only chance here is if someone like Apple abandons ARM for X86 for iOS but I doubt this will happen considering the resources they've dumped into their own custom ARM development. Microsoft has a compelling phone OS now but the public has pretty much written them off like they did with BlackBerry / BB10 so I don't see X86 making inroads there either.

Companies like Qualcomm in the west are doing are doing a great job cornering the tablet and smart phone market while companies like Mediatek is handling the developing countries.

No room for X86 in this game.
 

kimmel

Senior member
Mar 28, 2013
248
0
41
That's funny, I have numerous apps from the google play store on my Nexus 4 that don't run. Must be an ARM problem. :hmm:
 

TrulyUncouth

Senior member
Jul 16, 2013
213
0
76
That's funny, I have numerous apps from the google play store on my Nexus 4 that don't run. Must be an ARM problem. :hmm:

Just out of curiosity, what apps? My wife had the Nexus 4 and a strong appetite for apps but never ran into problems. What apps aren't working?
 

magomago

Lifer
Sep 28, 2002
10,973
14
76
I think he is making a joke that a lot of apps are pure garbage on the Android Marketplace...
 

NTMBK

Lifer
Nov 14, 2011
10,377
5,521
136
I think he is making a joke that a lot of apps are pure garbage on the Android Marketplace...

The article tried out the top 25 free and paid apps. You would hope that these would have some modicum of quality.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
You know what would be fun?

The Tablet tested is Samsung Galaxy Tab 3 10.1. I'm pretty sure somewhere out there has one.

Care to test top 100 apps. :D
 

kimmel

Senior member
Mar 28, 2013
248
0
41
Which ones? I also have a Nexus 4.

- Final Fantasy All The Bravest. The reviews were littered with don't download it doesn't work. I downloaded it. It didn't work.
- Star Wars Tiny Death Star crashed so much that I stopped playing it the same day I downloaded it. It was unusable.

There have been others but those are the two that jump to mind right now. You can say "they have been updated" and now work which is entirely besides the point. That also could be true for the apps that ARM is listing above on x86. We aren't given the list of apps that didn't work to validate their claims. I don't have an actual android x86 device to use to validate their claims either. heh.

The compatibility tests were listed for one single phone which just so happens to be probably the best case phone for compatibility. I'd love to see numbers on other phones. They aren't equal. Screen resolution matter to some apps, what version of processor matters. What versions of android matter. To blame this simply on ARM or x86 with a "run/will not run" criteria. Is quite the marketing claim.

The only thing this article tells me is that ARM actually cares enough to put out FUD articles about Intel now instead of just blowing them off as they did previously.

Also, as a side note the Chinese developer article that was linked is from May 2013. However it does contain a bit more details than the ARM PR piece.
 

kimmel

Senior member
Mar 28, 2013
248
0
41
I think he is making a joke that a lot of apps are pure garbage on the Android Marketplace...

Yes that was part of the joke. Yet that is the bar that is set by ARM marketing and the bar that people here are using to claim x86 has no place in mobile.

I look forward to having both ARM licensees and x86 fighting it out to drive prices down even more. There most certainly is room for both.
 

Exophase

Diamond Member
Apr 19, 2012
4,439
9
81
The performance degradation you get from binary translation is totally in line with a publication Intel made (that I cannot for the life of me find right now, sorry) that showed typical performance was < 50% native equivalent when not using a special instruction prefix to disable flags generation (that didn't make it into these Atoms).

That was a couple years ago, so they could have improved it since then, but you'll always have some overhead. Especially when going from 15 integer registers to 8, assuming they didn't find a way to execute x86-64 code in Android. If Intel really made a statement about this not affecting power at all they're totally taking the piss.

But I would have thought that the number of popular apps without x86 binaries would have gone way down since it has been a long time since it was first added to the NDK. I've co-launched an app w/o x86 support last year, since it relied heavily on ARM ASM for performance, but even we have since added in an x86 build..