Intel to give away 50,000 Arduino-friendly computers based on 400MHz Quark SoC

It's great to see Intel come out with a new microcontroller to compete with Cortex-M and -R series ARM devices in the embedded market, and to finally see its first high-profile platform in the form of Galileo.

When the Quark was introduced at Intel's Developer Forum 2013 a few weeks ago, the CEO seemed pretty keen to portray Intel as a leader in the wearables and Internet of Things  bandwagons (as well as in the tablet and the phone markets), despite everyone there presumably knowing that Intel is barely in those markets at all at present.

Maybe this is about to change, at least in wearables and IoT, since in this area microcontrollers can often suffice and application processors may not be needed.  Unfortunately Intel confuses the discussion here quite a bit, because in the opening paragraph (page 37) of their full 920-page Quark datasheet they call the Quark SoC X1000 an "application processor".  Oh dear.  I guess it's "just words", but it's never good to confuse discussions by redefining a term that has already acquired a stable worldwide de-facto meaning.

Being a conventional microcontroller, the Quark has no MMU to support the flexible process separation and virtual memory of full operating systems like Linux.  However, like the Cortex-M3, it does have the equivalent of ARM's Memory Protection Unit (MPU), which in Quark terminology provides Isolated Memory Regions (IMR) -- see page 127 of the datasheet.  Just like in Cortex-M3, eight different memory regions can be defined and there is a system/user separation of access privileges to provide a small degree of robustness against software faults.

The Arduino community into which Galileo is launched won't have any trouble understanding that Galileo is a bare metal microcontroller board just like most of their existing AVR-based equipment, and in that role it might do very well.  It's easy to forecast confusion in other communities though, especially those used to ARM Cortex-A* application processors running Linux distributions or Android.  A quick search already shows articles in the press suggesting that Quark is going to power tablets and phones, so the confusion has started.

Be that as it may, I think Quark might do well as a microcontroller, assuming that price and availability don't become a barrier.  The Galileo board seems unnecessarily big and complex for an IoT platform, let alone a wearable one, but I'm sure that tiny and cheap boards will follow.  The interesting times in embedded have just become even more interesting.

This thread on the Intel Quark SoC is quite interesting, but theres one big piece of Quark that has been overlooked.  Having worked with Intel architectures for some time now, I'm familiar with the way they write their manuals.  And what isn't there is always more important that what is!

Unfortunately Intel confuses the discussion here quite a bit, because in the opening paragraph (page 37) of their full 920-page Quark datasheet they call the Quark SoC X1000 an "application processor".

and

Being a conventional microcontroller, the Quark has no MMU to support the flexible process separation and virtual memory of full operating systems like Linux. 

Actually, Quark does have an MMU, its just not described in the Datasheet.

The MMU is part of the CPU Core, which is described the Core Hardware Reference Manual and the Core Developer's Manual.   The Quark CPU Core is a complete IA32 implementation,  a little slow but with a few extras for the embedded micro market.  Check out page 20 of the hardware ref, it shows a block diagram of the core.  The three manuals together make up the Quark documentation.  You almost need a monitor for the datasheet, the hardware ref manual printed out and a second monitor for the developers manual to make sense of it all.

The Galileo board seems unnecessarily big and complex for an IoT platform, let alone a wearable one, but I'm sure that tiny and cheap boards will follow.  The interesting times in embedded have just become even more interesting.

Very true..  I'd like to get my hands on one (or two?) of the Galileo boards myself.

Walt Gribben wrote:

 

Actually, Quark does have an MMU, its just not described in the Datasheet.

 

The MMU is part of the CPU Core, which is described the Core Hardware Reference Manual and the Core Developer's Manual.

Ahhhh!

Many thanks, Walt.  That makes all the difference in the world.  I guess Intel technical authors just have a warped sense of humour in not even mentioning the existence of an MMU as a bullet point in the SoC datasheet.  Perhaps there just wasn't room, they had only 920 pages after all ...

Cool, this puts a totally different perspective on Quark and Galileo, and yes, Quark is now a proper applications processor and not just a microcontroller.

Morgaine Dinova wrote:

 

Many thanks, Walt.  That makes all the difference in the world.  I guess Intel technical authors just have a warped sense of humour in not even mentioning the existence of an MMU as a bullet point in the SoC datasheet.  Perhaps there just wasn't room, they had only 920 pages after all ...

You have to view that datasheet much as you would the chipset hardware datasheet that intel would provide for something like the PCH as that's really what it is.  There's a single page (121) with cpu core details. Intel has traditionally done things this way, one doc covering the hardware side, another for the cpu internals/software/programming side.

Walt did well finding those manuals, I was coming up blank searching for them on the Intel site.

At 400Mhz though, it's difficult to know where it fits. Like the Minnowboard, it's too slow, too power hungry and will be easily outperformed by the various cheap Arm boards. So it'll come down to price.

Giving them away to universities is perhaps understandable, Intel is just trying to capture another generation in much the same way as MS does with very cheap versions of it's software.

selsinork wrote:

 

At 400Mhz though, it's difficult to know where it fits. Like the Minnowboard, it's too slow, too power hungry and will be easily outperformed by the various cheap Arm boards. So it'll come down to price.

Indeed.

At the  $60 to $50 prices being cited unofficially, I'm having a hard time seeing exactly what its selling point is going to be.  Some people will see the Quark's x86 instruction set as a selling point, but that's actually a complete delusion as benefits go since I don't think it means that modern i686-optimized binaries are going to run on this board.  Instead that old instruction set is quite probably a legacy burden on power efficiency and performance.

Add the AM3359-based Arduino TRE into the discussion and Galileo could be a very hard sell.  The more I read about the TRE's design the more I'm impressed at what TI/CircuitCo/Arduino have done.  But again, your words apply --- "it'll come down to price".

Morgaine Dinova wrote:

 

 

At the  $60 to $50 prices being cited unofficially, I'm having a hard time seeing exactly what its selling point is going to be. 

Yep. Make it less than $20 and it still only rates a 'maybe'.

Some people will see the Quark's x86 instruction set as a selling point, but that's actually a complete delusion as benefits go since I don't think it means that modern i686-optimized binaries are going to run on this board.  Instead that old instruction set is quite probably a legacy burden on power efficiency and performance.

Well it is a selling point, like it or not. It's hard to put an objective value on that warm fuzzy feeling of being on familiar territory.

i686 is really a fairly old architecture now. Not something I'd personally call modern.  In any case, someone will have to get their hands on one of these to find out exactly what cpu capabilities are present, what instructions are missing etc before it becomes clear how useful the processor is going to be. 

One of the interesting things I picked up while reading up on the SoC is that aparrently they're going to make the core available in synthesizable form for the first time. That alone probably tells us there will be limitations in one way or another.

Morgaine Dinova wrote:

 

 

At the  $60 to $50 prices being cited unofficially, I'm having a hard time seeing exactly what its selling point is going to be. 

Yep. Make it less than $20 and it still only rates a 'maybe'.

Some people will see the Quark's x86 instruction set as a selling point, but that's actually a complete delusion as benefits go since I don't think it means that modern i686-optimized binaries are going to run on this board.  Instead that old instruction set is quite probably a legacy burden on power efficiency and performance.

Well it is a selling point, like it or not. It's hard to put an objective value on that warm fuzzy feeling of being on familiar territory.

i686 is really a fairly old architecture now. Not something I'd personally call modern.  In any case, someone will have to get their hands on one of these to find out exactly what cpu capabilities are present, what instructions are missing etc before it becomes clear how useful the processor is going to be. 

One of the interesting things I picked up while reading up on the SoC is that aparrently they're going to make the core available in synthesizable form for the first time. That alone probably tells us there will be limitations in one way or another.