SoC: Back and forth!

[Grench]

I'm still a bit scared by the power consumption of the Intel... according to specs, it would be up to 6A@1V - which means means 6W (only for the core).

Add display, etc. around it, and it's quite a bit of power.

I looked around a bit and couldn't find any reference to 6W or 6A@1V on the z3770. I found reference to SDP 2W and max TDP 4W. There are other processors in the Bay Trail family that have much higher power needs - is there any chance you were looking at another SoC?

A TI OMAP5 developer mentioned to me, that one of the design guides they had for the OMAP5 was 2W maximum power usage.

Not sure if they really did that (we need to measure it), but 6W sounds like quite a bit, only for the core.

6W does sound like a lot - but if that is true, wouldn't it be 'peak' with all 4 cores tapped out? Does the SoC self-throttle under thermal protection - or would we be able to throttle it back and only -give- it a max 4W? I've seen the 4W number thrown around about the OMAP too - but we seem to have even less information about it's true thermal requirements.
There are a lot of myths and fears floating about here. I'm glad you're giving it consideration.

 

[Askarus]

Shouldn't this discussion be OMAP 5 vs Snapdragon 800/805?

Don't see any reason to use X86.

 

[Natsu]

Don't see any reason to use X86.

*Devil's advocate
It's stronger than the OMAP5 and people are asking for it.

 

[TrashyMG]

Going to x86 despite all the merits and such will fracture the existing community base, which I see as the most important aspect of this handheld project.

 

[_jr_]

IIRC the Pyra case and layout will be open. Wouldn't these significantly lower the cost to build a variant with a different SoC? 10,000$? I suspect ED would even be as helpful as possible for anyone trying that.

I personally don't like Intel. They are a business, but they've shown over and over again that they will do nearly anything in the name of profit where competitors in similar circumstances tried to play nice (and suffered for that). Let's see (these are obviously simplified): trying to shove down a binary blob WLAN (Linux) driver for a very popular platform down peoples throat not that long ago (using stupid excuses which were forgotten in a blink when pressure became high enough), while competitors - like RaLink - were already providing more open drivers. Taking taxpayers money and running (I think it was legal, the contracts were crap) after German reunification while competitors - like AMD - stayed and provided jobs (and did reasonably well for a time, but there were probably much better business opportunities, especially if they would have done the same stunt as intel). Using patents to stifle competition when suitable, but crying like a girl when someone else gets the upper hand in the stupid patent meta game. If a company is doing exceptionally well in terms of profit per employee it is most probably someone else that is paying the bill, it is nearly never a superior product. And if a thug uses his accumulated wealth to do some good it is still immoral to me.

 

[wej]

Don't see any reason to use X86.

How about open source drivers for the whole SoC?

 

[_jr_]

Don't see any reason to use X86.

*Devil's advocate

It's stronger than the OMAP5 and people are asking for it.

*Devil's advocate

But - excepting Exophase or _wb_ - not the important productive members of the community


edit: added _wb_ just to be on the safe side, if I forgot someone please shout

 

[Askarus]

I'd say if we have a chance to get the Snapdragon 805 it's superior to the Bay Trail.

 

[Hồng Thất Công]

I'm just drawing the x86 logo

 

[Grench]

For me, I have always wanted a mobile WoW handheld. With x86 that would be possible, along with mountains of other games, with ARM it will never be without source.

 x86 won't give us more games. Let me explain :

Just plain wrong. It opens up native Linux Steam games and native Linux commercial X86 Linux games (Ex: Neverwinter Nights). Pretty much anything that will run on a 64bit Debian installation on a 2Ghz Core2Duo should work similarly - except it's running in a 2W envelope. This opens up the entire Debian repository, Steam, etc...

Atom is already struggling with games, you can barely play on low details on a stripped down Windows.

The term "Atom" covers a long line of junky CPUs over many years. Don't confuse that with the z3770. Although it's being marketed under the "Atom" brand name, this is something entirely new - and realy freekin' cool. Nobody has said anything about running Windows on the result. A few people have put Steam on their Bay Trail tablets and gotten great results with older games - and meciocre results with 'cutting edge' games. Exactly what would be expected should they do the same thing on a Core2Duo with Windows 8 overhead strapped to it. We'll be running Debian Linux with XFCE - which should free up even more resources to the applications.

Here we are talking about an embedded device (the architecture will totally differs from a desktop or laptop PC, no choice)  which will not run Windows. It will be Linux. Period.

The Bay Trail SoC that we're talking about is for embedded and mobile devices. Yes, I agree that we're talking about a Linux machine - but that wouldn't stop someone from installing Windows on it on their own. Hell, people have installed Windows on the Pandora for the giggles and grins of it. I use Linux on many machines that have Windows stickers on their cases. Mobile devices do not 'totally differ' from desktops. In fact, the Pandora has much more in common with a chromebook laptop than it does to a cell phone. The SoC being discussed IS an embeded SoC designed for exactly what we're talking about.

The only advantage given by x86 when using software compatibility layers is avoiding the cpu emulation part. Is that really worth it ?

Not the only advantage. It appears that it might be faster in a similar or even smaller power envelope. It has open source drivers for everything - Including the graphics. The Intel HD graphics on it aren't spectacular compared to a dedicated graphics chip (Nvidia or other) on a laptop, but the stats so far show that it out performs most mobile and embedded SoCs.

People thinking that a underpowered desktop cpu trimmed down for mobility will handle games on top of a costly compatibility layer will be disappointed.

More fear mongering. This isn't a throttled down power hungry desktop CPU. It is a purpose built mobile and embedded SoC line that just happens to be X86. People are successfully playing older steam games on Bay Trail tablets. Unfortunately, most of those tablets appear to have junk boot loaders that make loading Linux on them difficult.

Someone should really explain the whole Wine thing (also Apkenv) in a detailled post, and it should be stickied like the "no source, not port" thread.

So far, to my knowledge, there have been no practical results from WINE through QEMU on ARM. Running native X86 Linux software on ARM appears to be a pipe dream with no real development going on.
All of your claims appear to be emotional rather than factual. Can we look at the facts and see what falls out?

The z3770 isn't your grandfather's X86.

 

[notaz]

(I think I'm repeating myself but whatever, I'll try not to post about this any more)

and I do think the people that are harping on about x86 are under an illusion that it's somehow the magic bullet for software...

Not at all. As I said earlier, the actual architecture does not matter in most cases anyway.

For you it may not, and in ideal world it shouldn't, but when you start to write more "interesting" software or you need to do some more serious debugging, it becomes all the difference in the world.

I fully agree that you should not trust the manufacturers and their numbers, but that is true for all of them. I still think that the Anandtech comparison is pretty objective.

I wouldn't be so sure. They said themselves that the measurement kits were provided by Intel, and who knows what are they measuring there. It could easily be the old "northbridge" trick where the core power usage is measured, but some vital support circuitry (for the CPU) is not. For me those articles are no more trustworthy than if they were from Intel themselves. That being said Intel surely has all the resources and fabs to build something that could outperform ARM, no doubt about it, so if there is any cheating, it's probably not very awfully significant.

--

In my opinion the architecture (or ISA to be more precise) is significant part of the device, it's kind of a "spirit" inside of it. When I think about devices, for example GCW Zero, first thing that comes to mind is not how it looks or the screen resolution, but the MIPS arch. Given that this community originated from GP32 -> GPH handhelds -> Pandora (as spiritual successor) which were all ARM, I would say the successor should stay ARM too.

It's up to ED to decide in the end.

 

[Grench]

Going to x86 despite all the merits and such will fracture the existing community base, which I see as the most important aspect of this handheld project.

Irrational fear mongering. Get your pitchforks and torches! The other is coming!

This community survived Craig. It can survive a change in SoC architecture that 99% of the time will be invisible to the end users.

Concerns about X86 bringing in a bunch of filthy non-ARM people is a big pile of elitist BS. If the worst that happens is we get rid of a few snobs that think this 'OPEN' device has to be closed off from the world - then we're better off without them. The Pyra would still be a nich Linux gaming oriented handheld device.

Lets look at facts, stats and capabilities. Compare them rationally, then pick what suits the envisioned device best.

If ED were taking pre-orders today, SoC undetermined, I would still be willing to jump in because I know it will be great. Allowing him the information needed to choose wisely and not promoting irrational fear would probably be a good thing.

 

[God Ginrai]

In my opinion the architecture (or ISA to be more precise) is significant part of the device, it's kind of a "spirit" inside of it. When I think about devices, for example GCW Zero, first thing that comes to mind is not how it looks or the screen resolution, but the MIPS arch. Given that this community originated from GP32 -> GPH handhelds -> Pandora (as spiritual successor) which were all ARM, I would say the successor should stay ARM too.

I couldn't have said it better myself.

EDIT:

Going to x86 despite all the merits and such will fracture the existing community base, which I see as the most important aspect of this handheld project.

Irrational fear mongering. Get your pitchforks and torches! The other is coming!

This community survived Craig. It can survive a change in SoC architecture that 99% of the time will be invisible to the end users.

Concerns about X86 bringing in a bunch of filthy non-ARM people is a big pile of elitist BS. If the worst that happens is we get rid of a few snobs that think this 'OPEN' device has to be closed off from the world - then we're better off without them. The Pyra would still be a nich Linux gaming oriented handheld device.

That's not what he means, and you know it. He's talking about how there is no community to speak of for other x86 devices. I don't frequent some special community for my Sony Vaio Laptop, and I'm sure you don't for any laptops you may own either.

The point is that if we go x86, users will treat it like any other x86 system and just mess around on it themselves. These forums would at best be a place where people only came for tech support problems.

-God Ginrai

 

[SNESFAN]

Here we are talking about an embedded device (the architecture will totally differs from a desktop or laptop PC, no choice)  which will not run Windows. It will be Linux. Period.

The Bay Trail SoC that we're talking about is for embedded and mobile devices. Yes, I agree that we're talking about a Linux machine - but that wouldn't stop someone from installing Windows on it on their own. Hell, people have installed Windows on the Pandora for the giggles and grins of it. I use Linux on many machines that have Windows stickers on their cases. Mobile devices do not 'totally differ' from desktops. In fact, the Pandora has much more in common with a chromebook laptop than it does to a cell phone. The SoC being discussed IS an embeded SoC designed for exactly what we're talking about.

it is my understanding that any OS that supports UEFI will boot on those bay trail chipset devices to include windows (as long as secure boot is off/disabled)

http://technet.microsoft.com/en-us/library/hh824898.aspx

Other OS's (OSX, legacy versions of windows etc) will need some sort of secondary boot loader to load the desired OS, sort of like windows PE or a minimal linux boot menu.

I was suggesting installing windows over stock, not supporting it stock, Thorgan seems pretty confident that it's not possible though.

 

[Natsu]

z3770 A more compelling reason for more ram?

 

[Hồng Thất Công]

--

In my opinion the architecture (or ISA to be more precise) is significant part of the device, it's kind of a "spirit" inside of it. When I think about devices, for example GCW Zero, first thing that comes to mind is not how it looks or the screen resolution, but the MIPS arch. Given that this community originated from GP32 -> GPH handhelds -> Pandora (as spiritual successor) which were all ARM, I would say the successor should stay ARM too.

http://boards.openpandora.org/topic/15560-soc-back-and-forth/?p=312417

No ducking this time then

 

[Wrath Of Khan]

Going to x86 despite all the merits and such will fracture the existing community base, which I see as the most important aspect of this handheld project.

Irrational fear mongering. Get your pitchforks and torches! The other is coming!


This community survived Craig. It can survive a change in SoC architecture that 99% of the time will be invisible to the end users.


Concerns about X86 bringing in a bunch of filthy non-ARM people is a big pile of elitist BS. If the worst that happens is we get rid of a few snobs that think this 'OPEN' device has to be closed off from the world - then we're better off without them. The Pyra would still be a nich Linux gaming oriented handheld device.


Lets look at facts, stats and capabilities. Compare them rationally, then pick what suits the envisioned device best.


If ED were taking pre-orders today, SoC undetermined, I would still be willing to jump in because I know it will be great. Allowing him the information needed to choose wisely and not promoting irrational fear would probably be a good thing.

My thoughs exactly!!

 

[Grench]

The z3770 has open source drivers.  It is supported  in the Linux kernel.  It meets 100% of the other requirements for an SoC for this system, but it is X86 not ARM.

 It is even available though Mouser in single unit quantities at around 40 EUR. They even have 100 in stock. But no data sheets I.e. we can't even check if and how we can connect anything.

What makes me nervous is that it is said to be a "FCBGA1380" in a 17x17mm package (OMAP5 is 17x17 mm PBGA754). This means it may need a more expensive pcb technology to be connected. And if I interpret some rough diagram correctly, it has only one eMMC and one SD card interface. This makes interfacing two SD slots plus internal eMMC more difficult (could go through USB and an extra chip) and WiFi as well. So the total system complexity must be taken into account.

But it is quite useless to buy chips without getting public documentation. For the OMAP5 there are 6000 pages of technical reference manual available.

At the moment it looks to be the first time that we have a small chance to get an up-to-date SoC with x86 architecture (it was impossible for the last 10 years). But if this is the right train to board is a different question.

Data sheet is linked to from the Intel product page:

http://ark.intel.com/products/76760/Intel-Atom-Processor-Z3770-2M-Cache-up-to-2_39-GHz

Right hand column. "Download Datasheet" It's not like they hid it.

I have no basis to argue whether the PCB technology is more or less expensive.

eMMC and SD interface overview is in section 1.2.11 of the datasheet.

Supports one SDIO 3.0 controller

Supports one eMMC 4.51 controller

Supports one SDXC controller

Personally I'd put the WiFi on a USB 2.0 port so that it's easier to swap out in the process if the WiFi chip chosed goes EOL. Isn't that easier and would have faster data?

Per page 21 of the datasheet, the Z3770 has One USB 3.0 OTG port and an internal hub with 4 USB 1.1 to 2.0 ports.

Anyway, the documentation doesn't get any more 'public' than posted in the open on the product page. There are 291 pages on V1 (overview) and there is a 23MB V2 as well - I'm pretty sure you'll find anything you need there IF the project is serious about considering this option.

I agree that it is possible that the Z3770 may not the right SoC for this project. However, Intel has seemingly pulled a huge rabit out of it's hat and seemingly created exactly what our specs call for.

It's fast, efficient, available, has open source drivers and is relatively innexpensive.

It may not be for us, but wow - we should at least consider it.

Of course there is something attractive to me about having a 64bit X86 Debian handheld running Half Life and Neverwinter Nights. So I'm biased.

Before someone piles on the 'We've been through this! It's Done! You're an idiot for bringing it up!'

I was directly addressed from the system designer with their concerns - and misconceptions surrounding this chip that they apparently had -not- considered. This reply is to allow them the option, if they choose to explore it, to see the information that was claimed to not exist.

 this is indeed the missing piece of information. Maybe it is not that prominently published as by TI or simply not that well known (meaning intensively linked to from the outside).

Many thanks for the link to the data sheets. They indeed look impressive.

Well, we *may* be able to offer different and exchangeable CPU boards if we can squeeze them all onto a small and flat CPU only board... Maybe in a second step... Just dreaming

Anyways it is EDs decision.

Thank you for giving it consideration. There are a lot of irrational claims and heated statements revolving around this topic - which you can see displayed here. Several of us are seeing your efforts as an independent, rational and clear mind to give us the reality of what is/isn't possible or even a good idea.

It's a relatively new SoC (months old not years) and the docs are conveniently listed on a public product page, so it isn't surprising that it's not re-posted all over the place.

The idea of exchangeable CPU boards has been floated about before. I think it's a great idea. No, it wouldn't 'fracture the community' but expand it. Room for more in the boat.

So, for the board, you'd be building something like this only with a z3770 instead of it's hungrier cousins?

http://www.portwell.com/productnews/PCOM-B632VG_PR.htm

http://linuxgizmos.com/industrial-strength-com-express-module-offers-bay-trail-socs/

http://linuxgizmos.com/intel-atom-e3800-socs-head-for-embedded-systems/

http://linuxgizmos.com/qseven-and-com-express-boards-run-intel-bay-trail-i-socs/

And yes, I agree, in the end it will be EDs decision. My efforts here are to ensure that he's considered all of the possibilities based on facts so that the end result is the 'best' handheld gaming and general use computer that it can be.

I look forward to seeing your updates. Thank you!

 

[God Ginrai]

The idea of exchangeable CPU boards has been floated about before. I think it's a great idea. No, it wouldn't 'fracture the community' but expand it. Room for more in the boat.

Speaking of "irrational claims"....

-God Ginrai

 

[hns]

I'm still a bit scared by the power consumption of the Intel... according to specs, it would be up to 6A@1V - which means means 6W (only for the core).


Add display, etc. around it, and it's quite a bit of power.

I looked around a bit and couldn't find any reference to 6W or 6A@1V on the z3770. I found reference to SDP 2W and max TDP 4W. There are other processors in the Bay Trail family that have much higher power needs - is there any chance you were looking at another SoC?

Data sheet Intel® AtomTM Processor Z36xxx and Z37xxx Series page 261 Table 118. Bay Trail-T SoC Power Rail DC Specs and Max Current:

VCC 6.5 A, VNN 5 A

And table 119 rates these voltages up to ca. 1 V.

Well, this is certainly the absolute maximum to design for (width of copper wire traces and DC/DC converters) and in practice it will be much lower. And it appears that it is limited by throttling the CPU if it becomes too hot.

Anyways it is really difficult to get a fair comparison.

Most likely other factors will outweight the pure peak performance or other features. The decision will certainly be based on more than 10 factors - and my impression is that every option has its pros and cons.

So throwing a dice could be the best solution h34r:  Then nobody can say that it was a bad decision because this or that was not considered