How do the Intel Quark series or the Curie series compare to the other 32 bit offerings in the industry notably ARM, PIC32 ...
Would there be any good reason to shift to Intel's offerings ....
My criteria for comparison is the following...
1. Good range of analog peripherals ADC, DAC, and other unique offerings that microchip has like CTMU, Analog comparators ...
2. Good range of digital peripherals like Timers, Input Capture, Output Compare, PWM...
3. USB, Ethernet... etc..
4. Decent computational power.
5. Decent power consumption ...
Would like to hear your views on these...
TIA
IMHO, by far and away the best reason NOT to choose an Intel processor is ready availability of data sheets, technical support and samples. It's nothing to do with the silicon itself.
Can you easily obtain data and parts? If so, then it makes sense to compare chips. If not, then personally I wouldn't waste any time trying; I'd go with an ARM device instead, depending (of course) on your definitions of "good", "decent" and so on.
I just checked but the Quarks are available from Farnell (Element14).
Intel's site is clearly not designed as regular microcontroller's site. You have to jump through like 4-5 pages to get a PDF (which they first open inside the page!). Also, their site is broken on Firefox, I had to use Chrome. Anyway knowing Intel I'm glad it's not under NDA.
From my initial investigation, they have 4 series:
- Quark SoC X1000
- Quark D1000
- Quark D2000
- Quark SE C1000
The X1000 is a SoC and actually a 400MHz computer. These chips should compete with the PIC32MZ (although it is a bit slower in processing power) and Cortex m7 chips, probably running some kind of embedded linux. Problem: they cost 50 euros each instead of the usual ~20euro BOM for a complete embedded computer. Upside: they are a small board you solder onto your product, which can make 50 euro seem cheap.
The D1000 seems like a cortex m0+ competitor with an proprietary/unknown CPU and complicated memory architecture (like 4 different memory regions for a small chip).
It has a 12-bit ADC. No DAC.
Only 2 timers, no PWM or capture to speak of.
Only 1 SPI master and 1 SPI slave.
Only 1 I2C.
2 UARTs
Sleep current down to 1.5uA which seems OK, wake up time 2.0us which seems OK. Voltage range down to 1.62V which seems low enough for direct battery operation.
The D2000 is more of the same. Again like like 4 different memory regions for such a simple device, same amount of peripherals but this time with 2 DMA channels, and a whopping 2 PWM channels. Don't known about capture pins.
Both of these series lack USB, ethernet, CAN.
All in all I think feature wise it's under offerings of competitor cortex m0+'s or the simplest PIC32s, and then you also have to go through much trouble to use them:
1) If you're already in the ARM or PIC regime, you have the toolchain, the debug cables and tools at hand. You probably need yet another Intel cable or devboard.
2) Chips available through limited set of suppliers, of which only Mouser is most accessible by most of us.
3) The chips are not particularly cheap. 4.21 euro in 1 qty? And still 2.73 euro in 4000 quantity? Go look up what that can buy you in ARM land. Much more, like STM32F3 series.
4) Upgrade paths? Little. If you outgrown a small m0+ or PIC32 you can upgrade to a bigger chip. With this you're stuck, need to change platforms, port all your peripheral code, etc.
Intel only offers the C1000 which is a much different beast. 32MHz x86 chip that does have a nicer amount of peripherals - like USB - but comes in a 144 pin BGA and costs 11 euro each in quantity (Ouch).
Not sure why anyone would use that, aside from the x86 selling point. The x1000 would seem most interesting to me.. basically an embedded x86 system with PCI-e 2.0 and DDR3.
You'd need a pretty good reason to go for a relative newcomer in this space when there are are already so many well-established options.
Seems Intel have been trying to get into embedded for decades and haven't really succeeded yet - I'd be concerned about long-term availability if it fails to get sufficient traction.
You'd need a pretty good reason to go for a relative newcomer in this space when there are are already so many well-established options.
Seems Intel have been trying to get into embedded for decades and haven't really succeeded yet - I'd be concerned about long-term availability if it fails to get sufficient traction.
Intel have had a number of successes in embedded processing over the last 4 decades, but they keep walking away from them. That's why its risky to choose Intel now. The 8048 had a strong market. Intel walked away early, and left that market to others. Intel walked away from the 8051 just as it was really starting to grow, and its still an important core. The 8096, 80196 and other more advanced embedded parts had considerable success for a time, and Intel walked away.
Seems Intel have been trying to get into embedded for decades and haven't really succeeded yet
You are forgetting Intel came up with the 8051 and before that the 8048. The latter was used in every PC and PC keyboard until USB came around! And lets not forget you can find x86/x88 microcontrollers in various electronics (I've seen them in hard drives for example). All in all they must have sold billions of their microcontrollers. IMHO it is more accurate to say Intel has lost the embedded market a couple of decades ago.
edit: As coppice noted: Intel seems to lack the patience to grow a new business. IIRC Intel acquired several companies in the past decades and either killed them or sold them off again because they didn't know what to do with them.
The D1000 seems like a cortex m0+ competitor with an proprietary/unknown CPU and complicated memory architecture (like 4 different memory regions for a small chip).
The core can best be described as a legacy-free x86. No different operating modes, linear memory, simplified instruction set. Also no MMU or FPU. The programmer's reference manual is available. The memory architecture is very similar to most other MCUs. Boot ROM, code flash, data flash, RAM and peripherals mapped into a single, linear address space.
The D2000 is more of the same.
The D2000 uses a different CPU (Pentium architecture, but doesn't have an FPU).
Seems Intel have been trying to get into embedded for decades and haven't really succeeded yet - I'd be concerned about long-term availability if it fails to get sufficient traction.
Intel at least used to take embedded systems and availability seriously. Eg. the 80186 was manufactured from 1982 to 2007. However I don't think even they know what their long-term plans are, and if they can't get a major design win for one of their current offerings, how long are they going to stick with it?
Its interesting to read all the gamut of answers ...
hans had said
The X1000 is a SoC and actually a 400MHz computer.
I'm amazed that Intel did eventually make a 400Mhz processor from what I read from the datasheets (i must admit I only read about the D2000) it is a
32Mhz processor.
Even
http://www.cnx-software.com/2015/11/06/intel-unveils-three-new-micro-controllers-for-iot-quark-d1000-quark-d2000-and-quark-se/ corroborates it.
Also Intel has recently launched a $15 board as
http://www.pcworld.com/article/3055372/internet-of-things/intel-on-the-cheap-chip-maker-ships-15-iot-developer-board.html says and then it has a tie up with Arduino and made a new board with them I don't remember the name off hand.
Peripheral wise the devices seem to be weak compared to the offerings of ARM / PIC etc.. but I think based on their website Intel is targeting IOT, automotive and wearable devices... Probably these sectors don't need those kind of peripherals.
The other interesting thing about some devices is a unique pattern matching engine that they have don't have much detail about this though...
The D2000 seems to be priced @ USD4 / pc on Mouser...
I agree that at the moment ARM / PIC are lightyears ahead in terms of MHz/MIPS/peripherals .... I just wonder why would Intel want to enter such a highly competitive sector with devices that don't stand up to the existing competition..
What would be the advantages of x86 architecture that these quarks have would it mean that I can easily reuse code written for the PC on to the quark?
There are 2 interesting features for the Intel Quark from their website
Pattern matching technology: This allows the microcontroller to learn through pattern recognition and differentiate appropriate response events. Having decision-making ability at the edge can provide a real-time response while being more energy efficient.
Internal sensor subsystem: This is a hardware acceleration engine with the ability to manage multiple sensors, allowing more peripherals to be supported. The sensor subsystem can handle wake events while the main core is in sleep mode, resulting in very low-energy consumption.
However the SE C1000 runs at USD14 /pc at Mouser.
at the moment ARM / PIC are lightyears ahead in terms of MHz/MIPS/peripherals ....
I doubt that to be the case. Even the newer CMx chips aren't that much faster in terms of MIPS/MHz vs. 8051.
Plus, the real comparison should be getting a job done within a given time frame. Being a CISC chip has its advantages there.
I just wonder why would Intel want to enter such a highly competitive sector with devices that don't stand up to the existing competition..
Only Intel can answer that. I think they are hedging their bets for the future. Also, going into ARM's territory can help deter ARM from going into Intel's territory. An offence that's really a defensive play.
The sensor subsystem is a Synopsys ARC core, but I don't know how it's exposed (ie. can you run your own code on it, or are you limited to prepackaged blocks supplied by Intel).
I doubt that to be the case. Even the newer CMx chips aren't that much faster in terms of MIPS/MHz vs. 8051.
Yeah, I should have been clearer there I was more trying to compare peripherals in this case eg. peripherals on on ARM cortex / PIC32MX vs Intel quark D2000
Perhaps no USB, Ethernet etc. no Floating Point Hardware support like the STM32F4 would be a reason to worry about for anyone wanting to make the switch.
Here is a good review of the D2000 quark microcontroller from another forum by user JMG
http://forums.parallax.com/discussion/comment/1371854/#Comment_1371854Collected features summary, with spec details
* CPU: Intel® Quark™ microcontroller D2000 SoC 32 MHz ( 8mA typ @ 32MHz, 272uA 4MHz Halt, 3.4uA RTC )
* 32 KB flash memory (internal)
* 8 KB OTP flash (internal)
* 4 KB OTP Data flash (internal)
* 8 KB SRAM (internal)
* 1x I2C (Master/Slave) to 1Mb/s, 7b & 10b Addr, 16 Deep FIFO Rx & Tx, DMA
* 1x SPI master supports up to 4 CS, CLK <= 16MHz, 4~32b frame, 16B (32B?) FIFOs, DMA, 16b Frame Counter
* 1x SPI slave, as above, but 3.2MHz Max.
* 2x UART, 5~9b, 16550, DMA + 16B FIFO, supports 9-bit addressing mode & RS485 300bd~2Mbd < 2% error, Fractional Baud
Hardware Handshake Interface supports DMA
* ADC 2.28 MSps SAR 19ch 6/8/10/12-bit (12-bit at 2.28 MSps and 6-bit at 4 MSps).
* 19 Comparator, Rail to rail input, Choice of (<600nA, .5us or <9.8uA, < 250ns), VREF (0.95V +/- 10%) or PinRef.
* 2x 32b timers, as either
PWM 32b HI, 32b LO set, or 32b Timers - No Capture or External Clock modes ?
* 25 GPIOs, INTs + Debounce on all pins. 12mA & 16mA drive, VOL spec, no VOH drive spec ? (but does have DDR)
* Real-time clock, 32.768kHz Xtal, 32b timer, 32B RAM battery backed
* Watchdog timer (seems to CLK from only 32MHz ?)
* 6x6 mm 40-pin QFN
* Regulator 1.8V Switchmode, but needs 47uH, 47uF, 470nF components.
* Clocks : 32.768 RTC kHz 32MHz Xtal or (RC Osc 2% after trim ?, says 10 bit trim code from OTP)
32MHz Xtal Osc trims 5.55 pF to 15.03 pF, in 16 steps
* Price : Mouser stock, 250+ $2.53
Summary:
As a MCU, ignoring the core, just looking at Clocks+Peripherals, I'd rate :
Oscillators : Quite good, but watchdog from 32MHz is surely an oops. Sparse specs on RC osc modes.
ADC: Quite good
Comparator: Good choice of power, and on 19 pins.
Serial : Quite good, with FIFO and DMA and reasonable Speeds. SPI has 4-32b & frame counter.
Unclear if this can support i2s ?
Timers: They are 32b, but options are Quite Poor, just 2, as PWM or Timers, no capture or External Clock modes ?
FIFOs : Muddled info on depths, various claims of 8,16,32B ?
CPU: Limited info exposed - eg Mul/ Divide - how many bits and cycles ?
Addit:
Confusingly, there is also a D1000, and a 'D2000 series'. The 'D2000 series' are quite different, and the D1000 seems almost exactly the same as D2000 - almost like D1000 is rev 1 silicon. Price is the same.
In the Data mostly small diffs show up, like
Feature : D1000 , D2000
SPI frames : 4~16b 8w fifo, 4~32b 32w fifo ?
i2c Clock : 400k, 8B fifo, 1M, 16B fifo ?
PWM modes : ? none , 2 Timers can do PWM
IO : 24io , 25io
RS485/HWfc: ?? not found , Has hardware Flow control.
ROM : 8k , No mention (has 8k OTP?)
Mouser Stk@Price: 4311@2.53 , 3695@2.53
Opcodes:
Hard to find D2000 info, still early ?
However, there is D1000 info here, which could give some guidelines :
http://www.intel.com/content/dam/www/public/us/en/documents/manuals/quark-d1000-programmer-reference-manual.pdf
which indicates
* opcodes efficiently support 8b and 32b, and 16b are there, at a prefix byte size cost.
* DIV,IDIV seems to be 64b/32b -> q32:r32, and IMUL,MUL looks to be 32b*32b -> 64b
* Branches support rel8, rel32 for shorter opcode size
* Usual mix of Add/Sub/And/Or/Xor/Test/Push/Pop/Call/Ret found in most MCUs
It would be interesting to know if there are any figures available regarding who is the largest selling vendor of 32bit microcontrollers.
If i'm correct probably Microchip leads the 8bit microcontroller sales.
who is the largest selling vendor of 32bit microcontrollers.
Market share data is widely available.
The trick is to define "vendor" and "32bit microcontrollers". Generally speaking, ARM wins by a large shot. The usual suspects, LPC, TI, and ST are players in the ARM market. Outside of that, Renesas and Freescale are huge.
It would be interesting to know if there are any figures available regarding who is the largest selling vendor of 32bit microcontrollers.
If i'm correct probably Microchip leads the 8bit microcontroller sales.
You could have taken 10 seconds to Google "global MCU market shares" and found that even now that Atmel's share has been added to Microchip's, they are well behind the market leader for 8 bit MCUs and behind for MCUs in total.
they have open datasheet, written in Intel's usual cryptic but comprehensive way. If you have the time to read through tons of pages of boring text, then it is there for you.
Ok; I'll bite. Where?
I haven't found any register-level documentation for the Curie module in the year since I bought my first "Arduino 101" (that's the Arduino board with the Curie module that people have mentioned. At $30 for a BTLE-enabled Arduino board, it's not a bad deal.)
I've found the marketing-heavy "fact sheet", and I've found the (loosely speaking) "data sheet" that has physical specifications, but nothing on how to program the beast, register descriptions, the ARC IOP core (or how it intereacts with the x86), or other features.
(Arduino sketches apparently run on the ARC core, in conjunction with some "OS" code on the x86. The OS code has been publised...)
I haven't found any register-level documentation for the Curie module in the year since I bought my first "Arduino 101" (that's the Arduino board with the Curie module that people have mentioned. At $30 for a BTLE-enabled Arduino board, it's not a bad deal.)
If I remember correctly the
Intel Curie module features the
Intel Quark SE C1000 microcontroller have you checked out their website for the relevant docs, the datasheet of that microcontroller should be available.
the Intel Curie module features the Intel Quark SE C1000 microcontroller have you checked out their website for the relevant docs
Ah; yes, that is helpful. It doesn't say much of anything about the ARC core, though.
Farnell carries them and the have the full datasheet (950 pages of pure fun) attached
Wish there was a way to bookmark certain answers these gems would soon get lost and other's will be unable to find it.
While back a company decided to build a board based on the AMD Geode, because it was new and stuff. The Geode became obsolete before the board design was finished.
I would never touch Intel, AMD, or the newcomer Samsung, for embedded system, for any serious business.