SoC: Back and forth!


As ED said it's ARM and some ppl continue to argue endlessly, they are spammers.

Want to discuss about that (joke) ?
 
I'm pretty sure the 805 is the fastest and 800 is still faster than the z3770.
[QUOTE='Askarus]Right
your gut feeling? or is the marketing machinery talking? I do not claim the z3770 is faster. Compering them in terms of processing power is very difficult! One thing is sure, all three of them are very fast!
You didn't but T4b above claimed he's sure the z3770 would be faster than the 805.

I also remember reading somewhere that the z3770 is somewhere around the 800 (as much as it can be compared), so the 805 will probably be faster (until Intel releases the next SoC and so on...)

[/QUOTE]I wouldn't put it that way. It's what I thought was true and I put it on that wiki page just because it came to my mind at that moment. I still think a comparison table on the wiki with citations and everything is the way to go.
 
There is unconstructive spam here which is overdone imo.

No offence but then there is a definite hate for x86 here that seems more emotional than logical.

I'm sure Ed won't pick x86 anyways.

Omap 5 will be fine but if given the chance there is no harm in considering snapdragon or other possible candidates.

Perhaps one of these other socs will yield some unique benefits.

Lets wait and see and keep an open mind.
 
Many others from both sides do as well.
[QUOTE='Strigoi]
[citation needed] all i'm seeing are fanboy arguments + ED's annoyance level rising because of them.

[/QUOTE]seriously?  -.- 

you don't see any benefit in... for example good quality open source GPU drivers provided by Intel? o_O

I also remember reading somewhere that the z3770 is somewhere around the 800 (as much as it can be compared), so the 805 will probably be faster (until Intel releases the next SoC and so on...)
... which doesn't matter because waiting for the 805 would cost us probably about 6 months which is no option (if I remember correctly). So at least we can assume we have a draw CPU wise.

[QUOTE='Askarus]
I know.

Wasn't the best thing I could say (also @ t4b).

Will try to be more friendly now :) .

[/QUOTE]no prob, everything is cool ; )
 
Because there are lots of different Android devices, with lots of different CPUs with different features. It is quite a different situation there.
Dalvik doesn't care. One device could be compiled hard float, another soft float, a third hardsoft, and apps would have just worked, that was the whole point of dalvik. The decision to keep it consistently hardsoft was made years ago because that's the way it was at the time.And you can't talk about native code because that wasn't added until recently, Android was hardsoft for years before native development was even a consideration. You'll need to do better than that.
 
Wasn't that too long to wait for 805 was for the products?
 
Last edited by a moderator:
Because there are lots of different Android devices, with lots of different CPUs with different features. It is quite a different situation there.
Dalvik doesn't care. One device could be compiled hard float, another soft float, a third hardsoft, and apps would have just worked, that was the whole point of dalvik. The decision to keep it consistently hardsoft was made years ago because that's the way it was at the time.
And you can't talk about native code because that wasn't added until recently, Android was hardsoft for years before native development was even a consideration. You'll need to do better than that.
I don't, because it does not matter what the Android guys did for which reasons. Also, even though native code wasn't supported right from the beginning on Android, they may very well have considered to include it at some point from the very beginning. Maybe their choice was arbitrary or their initial device did not support hardfloat and they stuck with it. I don't know. The situation with the Pandora is still different. There was only one device with a CPU with hardfloat support, so it would very well have been possible to use hardfloat, if there would not have been that PowerVR driver.
 
Last edited by a moderator:
I'm not too sure I'd call the open source intel drivers "good quality", at least not the one for the HD4000 chipset I'm using. They're better than AMD propriatary drivers, but a joke compared to nVidia propriatary drivers.
 
I'm not too sure I'd call the open source intel drivers "good quality", at least not the one for the HD4000 chipset I'm using. They're better than AMD propriatary drivers, but a joke compared to nVidia propriatary drivers.
How so? I've got basically the same GPU (HD2500) and it works very well for me, including hardware accelerated video playback. Compared to the proprietary drivers of Nvidia and AMD, Intel's driver lacks behind with the latest OpenGL versions, but other than that I really can't complain. Works very reliably here. Also, compared to GLES even OpenGL 3 is a big step forward.
 
X86 mobile on anything less than laptops have flopped _every time_ Exception to battery life being clamshells 15 years ago
The X86 crowd is forever projecting their beliefs onto forthcoming products, which turn out to flop.
 Its just intel subsidizing their low power line to stay in business on general computing, so far resulting in zero fruitful products, and even less so anything that runs linux

The argument about open source is null and void once you take into consideration on intel you need uefi bios, for which amd disclose a lot more specs than intel

Anyone remember the EXA driver for atom debacle, it took _YEARS_ before that performed well
Not to mention it was all the same, subsidizing netbook products to get more business, in turn killing most of them
And while we are at it, why did intel impose artificial memory restrictions??? Because they want to protect their cashcow business. Intel is still intel

The argument about OMAP5 being harder in that its less standard, is to keep others from copying wholesale, in the sense that everyman and his cow knows to set up an regular PC platform.

Its about know-how and inhouse skills. That is to a large extent what this community is built around. Hacking to make things better, perform well, do things they didnt at the time it launched.
 
Last edited by a moderator:
OK I read the review. It says:

The intel cpu has far more cpu power. The only thing were it is slower is Dalvik Java, and Intel says it will try to optimize it (there seem to problems with dispatching work to several cores). In native code the intel chip has 21k vs less then 18k iterrations per sec. Means it is around 15% faster. In all benchmark but Java and a Browser Benchmark ist aroun 15-20% faster.

The problem is, that the gpu is way weaker then the snap dragon. Around 30%. I am not sure what we want to do (emulation bottlenecks are most times the cpu not the gpu). But for games like Doom3, the Dark Mod, etc a faster gpu would be better.

http://www.anandtech.com/show/7314/intel-baytrail-preview-intel-atom-z3770-tested/3
 
I'm not too sure I'd call the open source intel drivers "good quality", at least not the one for the HD4000 chipset I'm using. They're better than AMD propriatary drivers, but a joke compared to nVidia propriatary drivers.
How so? I've got basically the same GPU (HD2500) and it works very well for me, including hardware accelerated video playback. Compared to the proprietary drivers of Nvidia and AMD, Intel's driver lacks behind with the latest OpenGL versions, but other than that I really can't complain. Works very reliably here. Also, compared to GLES even OpenGL 3 is a big step forward.
I get crashes quite a lot. Doing anything in kicad is pretty much synonymous with a force-restart of Xorg when you least expect it.

And just today, I got another complete force-shutdown (as if someone just flipped a powerswitch) from the drivers.
 
[QUOTE='slaeshjag]
I'm not too sure I'd call the open source intel drivers "good quality", at least not the one for the HD4000 chipset I'm using. They're better than AMD propriatary drivers, but a joke compared to nVidia propriatary drivers.

[/QUOTE]I've heard only good so far. I'm verry impressed and happy with my Intel HD graphics working in my 250$ Acer c720

Accoring to Phoronix the Intel 4600 open source driver is close to the performance of the propritars Windows 8 driver. Yes I'd call that a good open source driver: http://www.phoronix.com/scan.php?page=article&item=intel_windows81_ubuntu&num=2

You can't compare a full fledged Nvidia graphics card with a GPU working in a SoC. By the way among Intel and AMD, Nvideas open source driver (niveau) is the worst (due to complete lag of support and documentation up until recently)

seriously? -.- you don't see any benefit in... for example good quality open source GPU drivers provided by Intel? o_O
[QUOTE='Strigoi]
Apparently not.

[/QUOTE]I think you might be right. Now I also feel like "all i'm seeing are fanboy arguments" *rofl* ; p
 
Last edited by a moderator:
I get crashes quite a lot. Doing anything in kicad is pretty much synonymous with a force-restart of Xorg when you least expect it.
Strange, it never crashed on me in Kicad (or any other application). Are you using the latest driver version?
 
Not sure about driver version, I'm using the kernel version/Xorg version combination that does not result in Xorg being permanently pegged at 100% CPU usage.

It's especially common to crash in the PCB-editor when you zoom in/out.
 
That's funny how some people label all the ideas they don't like as "spam".

If I read this thread again:

- ED initially made a report about his choice of a SoC

- The Intel SoC was then reminded as a possibly interesting option

- In the light of some technical information provided by some people, even ED admitted it might be an option worth studying, at least

- Since then, some people who'd like an X86 architecture continued providing interesting data and facts, while on the other hand ARM supporters had basically the "x86 is evil", "here goes the merry-go-round" and "it has been decided, so let's not change our mind" poor arguments.

I think that if going the x86 route was utter nonsense, ED and the team would have discarded the idea very easily, either for technical or business reasons. If it's still under consideration, that's certainly because they are not narrow-minded, and try to have a pragmatic approach. As ED said, now it's all about testing and seeing which option is the best. Now the only constructive thing that can be done is to bring even more technical arguments, so an educated choice can be made. Personnaly I don't really care which SoC will be chosen, but at least I'd like to know *why* (in other words, I'd like to hear a more interesting reply than the fact it can bring some fanboys out of their comfort zone).
 
Last edited by a moderator:
Back
Top