Intel's homage to Raspberry Pi: The much pricier Minnowboard The huge popularity of the British-designed Raspberry Pi has caught Chipzilla's attention, and so you can now buy a similar bare-bones x86 PC named Minnowboard with a similar caseless design running an Angstrom Linux build. minnowboard Intel's Raspberry rival The system runs a 1GHz Atom E640 processor with integrated …
Dunno about back doors, but I'd be more concerned about having to deep throat Microsoft just to be "allowed" access to my own goddamned property.
> Ssshhhhh you're not allowed to point out facts, it gets in the way of the 'tards hysterical rants...
UEFI brings NOTHING to the table here. What few real limitations are present in legacy PC firmware are negated by things like SSD and MicroSD and just partitioning your disk.
On the other hand, buggy UEFI implementations (like Samsung) have a tendency to brick themselves.
Even if it's back door free and you can disable "Secure" Boot, it's still very likely it has tons of security critical bugs in there. UEFI is _huge_ compared to the BIOS, or to OpenFirmware which has the same functionality for a tiny fraction of the code. Although UEFI is in theory open source, it's hard to debug it since if you make a goof your hardware will likely be bricked.
So please Intel, stop riding a dead horse. UEFI is not going to rid you of all your legacy stuff (Windows wouldn't run then anymore) and soon it'll turn into it's own much bigger legacy.
Intel are just as out of touch with consumers as Microsoft. Watching them desperately flail around with their olde world paradigms is too embarrassing to even be funny, it's just incredibly sad.
Less power, Intel, less, and lower prices while you're at it.
Will they ever learn? Probably not. It's hard teaching old dinosaurs new tricks.
Not sure if they are clueless. I still need to dig into the specs and do a lot of comparing, but it looks like the thing can be hijacked into some applications where the Pi is not powerful enough and the "Commercial Options" makes the price tag...interesting.. even if you got to do a lot of developing/soldering( yes, that forgotten art still exists..) yourself.
Maybe El Reg could get their incrowd to do a Pi for Slice comparison?
Exactly, the price point is insane. Atom boards and CPU (and AMD E's) can be had for 50-70 dollars. Yes they are a little larger, need ram adding and probably use a bit more power but 200 bucks is silly. If it was 50-70 it would probably sell, maybe even quite well. If it has usb 3 and gigabit ethernet it could probably be a decent 'dumb NAS' (i.e. no parity raid) along with streaming some media assuming no transcoding. For that money I could just buy a netbook, hell I can wait till black friday and buy a laptop for 200 bucks, with windows and much higher specs.
There probably is value in it being fully open. There also probably is some value in it being x86 (a lot perhaps) but it is all dependent on the price. People found a way of forcing the pi to do a lot because it was crazy cheap, almost buying one on a whim and pushing themselves to find out what they could do. Whilst the minnow is probably benefits from a higher IPC and should have far better IO, it costs a lot comparatively so they really are limiting its sales and audience with that.
The key will be how blob free it is :)
As a device aimed at embedded developers it's probably decent, as an attempt on the kingdom of pi it is lacking, primarily in the price department. It would be interesting to see what they could do if they started a non profit to develop something like this and didn't try and milk it for cash. Given the amount they spend on advertising and brand development it might pay off. If they could do it for 60 bucks and create a large supporter base it could do well for them, albeit within a specific set of people.
minnowboard.org is in the process of becoming a non-profit foundation. although Intel has provided a lot of assistance for the MinnowBoard, all of the research, development, and production costs were done by Circuitco. Circuitco is also the manufacturer and board member of the non-profit beagleboard.org foundation.
This is stupid, Lets put everything from a standard atom setup on a redesigned board with Less expandability and double the price.
I want what they are smoking, must be some amazing stuff if makes you this stupid.
What they don't seem to get is that its not really the hardware that is making it really move, its the price. For as little as you pay for a Pi you can risk damage and destruction to the system even giving it to very small child, and not feel it in your pocket book so much when the thing goes tits up in a cloud of magic black smoke.
The problem with netbooks was precisely the opposite - that we had a great idea in 2007-2008, then it basically never changed, so in 2010 onwards we still had mostly 2008 spec machines selling for the same price. I'd love to have a "bolted on" upgraded netbook with more RAM, new processors and higher resolutions...
The "magic" variety is typically blue, although admittedly it's entirely possible Intel might have resorted to practising black magic instead, in a desperate bid to get us to buy this junk.
Yes, this product is obviously not aimed at the RPi buyers and seems to be intended for the "big boy" embedded systems manufacturers (who will no doubt get some juicy discounts) who produce products for corporations, not hobbyists on a budget.
For them, the Pi is a nice starter board and other products like the BBB (also made by CircuitCo I believe), or the Olimex Linux boards (some with wifi or SATA on board) are much more capable as Linux hosts and a great deal cheaper.
"intended for the "big boy" embedded systems manufacturers (who will no doubt get some juicy discounts) who produce products for corporations, "
Given the price...
Why wouldn't potential customers choose an existing iTX or similar x86, if they want Windows?
Why would potential customers be interested in x86, if they're not tied to Windows?
BBB == BeagleBone Black, right?
In a market where Windows doesn't much matter to the customers, x86 doesn't much matter either.
BeagleBone (and similar) could be a much more relevant starting point for low volume builds than the product in this article. For high volume stuff, there are other possible options (and x86 still won't be relevant). Low price custom SoC could be relevant. Intel don't do low price and Intel don't do custom SoC.
Odd.
> Why would potential customers be interested in x86, if they're not tied to Windows?
Well that's the key, isn't it? Say you're an OEM making annunciators for trains, or portable ATMs, or things that trundle round warehouses following lines on the floor. If your *current* products are based on WinTel technology and you want better, faster, cheaper, more reliable, more power and a board that hasn't recently gone out of production then yes: this could be for you.
If you're upgrading an existing product or giving it a mid-life boost you don't want to spend vast amounts of developer time (and time to market) redeveloping stuff, checking if drivers will still work and integrating peripherals, USB thingies and so forth. You just want to drop in a new "heart" to the machine and get it out the door.
However if you're building a mass-market product where space, power consumption and unit-price are more important and you're writing all the software from scratch anyway, then you probably wouldn't consider this board.
x86 is useful at the low end too. I have a Mac G5 but I can't run BSD because some stuff (mythtv) doesn't compile on a G5. x86 would be hassle free for BSD. Debian is fine, but I wanted to play with BSD too.
I thought Minnow would be more of a server to the Pi's client, but at 200 is way too much and one sata port? What's the point of that? Far cheaper and better to re-purpose a core2 system.
I'm still not sure why a GPU (even a rubbish one) can't be used for the RAID 5 parity checks and the SATA logic be brought on-chip.
It does appear to be aimed at more of the commercial embedded market, but if you wanted millions of the devices, surely you'd pick ARM or mips and save yourself a bundle.
A while back I made a list of the sorts of things you could get with around €250--275. It included:
* a PS3 with free game
* a Nexus 7 (which has since been upgraded slightly)
* an ARM chromebook
* An eMMC-based ODROID-X2 (with plenty of change to spare)
* 2x microSD-based based ODROID-U2s
* 4x Raspberry Pi Model B (or 3x plus a network/USB hub)
* 5x Model A Pis (a rough guess, though adding wireless cards might push me over budget)
* Various combinations of {Pi, Arduino} and {gertboard, Pi Face, Slice of Pi, Adafruit, Arduino modules} and {basic electronics kit and tools}
Since then I see that the Parallella boards are available for pre-order, so I could add:
* 2x Parallella boards (with 16-core coprocessor and FPGA)
For what it appears to be (a hobby or "gadget" item), the price is just crazy. The only thing that it has that the other things above don't is PCI Express and SATA (the PS3 and Parallella have gigabit Ethernet and the Chromebook has 802.11n, so fast networking isn't unique to this board). Is that enough to warrant paying twice (or more) the price of most of the other things I listed? I seriously doubt it...
I notice that all the gagets I mentioned (bar the PS3) happen to be ARM-based, so perhaps that shows a bias on my part. On the other hand, it shows the range of products that Intel is competing against in this segment of the market--let's call it the "gadget" segment. As such, this new board would be at the bottom of my list, even assuming it made the list at all.
Parallella, eh? Sounds interesting. Come a long way from the Transputer eval boards, yes we have.
I want a few.
WHAT ???!!!
I can buy a complete computer for $199 AND it's expandable AND it can be converted to run Linux.
Here are the specs:
Screen
11.6" (1366 x 768) HD display
High-brightness, 200 nit screen
HD Webcam
Size
285 x 202 x 27.35 mm
Weight
3 lbs / 1.38 kg
CPU
Dual-core Intel® Celeron® Processor
Ports
3 x USB 2.0
1 x HDMI
1 x VGA
2-in-1 card reader supporting: SD, MMC
Memory
2 GB DDR3 RAM
16 GB Solid State Drive*
Audio
Stereo internal speaker
Headphone/microphone jack
Battery
4 cell, 37 Wh battery. Up to 4 hours*
******************************************************************************
Oh, by the way, it's called an Acer Chromebook C7M!
Take your Minnowboard and stuff it, Intel!
I must say I'm underwhelmed. The brilliance of Raspberry Pi isn't performance so much as its the ability to throw a huge codebase at relatively simple tasks with very little development effort. Networking, big displays via HDMI, perl, html, etc, cheap and easy.
Many years ago I was involved in a project which used FreeBSD on a Soekris board. 486 compatible at 133 MHz or so running off compact flash card. Needed a sound card and had trouble finding one that ran off 3.3V PCI. A Raspberry Pi would have been just the thing.
I agree, yet intel really do need a competitor. The pi is training the next generation of embedded engineers and on intels biggest competitor.
Intel need to get their game together and offer a similar product at a competitive price, however i feel the sales guys would fear it would undercut existing products so reducing the high margins, without thinking about the long term effect.
For hobbyists and embedded systems people?
Hobbyists don't need either the power nor the IO of this board and will not pay USD 200 just to have something to tinker with. But not I'm not even sure that the embedded developers would find it an attractive proposition: that heat sink and the rods indicate that it needs quite a bit of clearance and cooling. In fact the whole thing is the size of a Mac Mini which was impressive when it came out, what seven years ago, but has since been successfully copied and improved upon.
That said, I think the real point of this product may be in open sourcing all the hardware design. It can't have been easy to get past some managers and does probably open the doors for similar but less powerful boards costing significantly.
Having been an embedded developer for 30 years, I would say that what people **might** use this for is as a reference design.
If you just wanted to put a bunch of Atom boards in boxes (as you might with a RPi, ATX form factor board or whatever), then there are cheaper options.
The minnow board is pretty cheap when compared to its direct competitors: reference designs with schematics etc. The cheapest I could find elsewhere is around $500.
Hmmm...
My current robotics project uses an Atom based Mini ITX board with BeagleBone black boards to run functions for sensing and actuators on an SPi bus.
The whole lot including disk and memory cost less than this board and from what I can see there is nothing unique or special about it. If it is to capture the hobbyist market then it needs to be around the $60 mark.
ARM chips run very fast, BeagleBones are cheap as chips and have a lot more IO than a Pi, though I do have a model B sat in the workshop ready for some serious tinkering. If Intel want to send me one then I'm happy to run it through its paces and see what the differences are ;)
Its possible to get Arm boards with built in Gyros and Accelerometers, plenty of IO and a quick enough chip to run in a stack for about £25 each... Pi doesn't really have enough IO. And yes, you do need a lot of IO for robotics...
Intel have long supported open design, The first Pentium Pro servers from IBM were Intel boxes sprayed blue and one of the best engineered desktops was a 486 Intel OEM unit that sold for about half the cost of the components from other manufacturers. Memories eh?
In my Ham Radio days, there was something very satisfying about communicating with someone on the other side of the world with 3 watts of Morse (HW8, anyone), vs. the 'Big Boys' - Stateside - with 1 kilowatt, huge steerable antennas, and all the DSP they could afford to throw at it.
It's a parallel to what can be done with a Pi. Pi IS limited, but the ingenuity of smart folks relish in the challenges it sets. Case in point being "FishPi". Sure a nice multi - K$ can do it, but a $35 computer???
Brings me back to the days I'd sweat over code in the wee hours, just to get my program down from 1027 bytes (ROM was pricey then, folks!).
Anyway, as pointed out, this box is waaay overpriced, and seems to have suffered from a bout of 'feature-creep' from initial idea - if there indeed was one.
FishPi: http://fishpi.org/
How appropriate for an obese man to present this joke.
This is massively overpriced for small projects, like all Intel CPUs and mobos, and 1GB is very stingy after virtual extensions support was throw in!
Even a small server can need much better I/O support, even for this form factor. I can buy AMD Fusion embedded boards for a fraction of the price, which are significantly better than /all/ of the off-the-shelf embedded Intel Atom boards I am aware of! I seriously question the sense of people who build NAS and servers with Atoms, yes especially off-the-shelf SME boxes.
AMD is probably going to murder Intels' embedded market share when their SoC Fusion APUs are rolled out.
Intel /really/ need to significantly drop their prices and provide /much/ better I/O support, as standard, on it's embedded boards, because having to buy extra cards just to have practical base I/O support, takes the P!
If this really is intended to compete with RaspberryPi the name is a complete fail.
To be hip and hyped you need stupid name like RaspberryPi, Jelly bean, Ice cream sandwich?
A name where most people would have no idea what you are talking about so just knowing what the name is makes you hip.
Minnowboard - it is a board and kinda like a small weak fish - They stand more chance of selling them if it were called 'Knickerbocker Fanfare'. The price is also stupid but I guess they know they will never sell the volume required to bring it to sub $50 level.
That's because it's not intended to compete with Raspberry Pi. From the name, price point, form factor, and level of embedded I/O, it's intended to compete with TI's BeagleBoard and its family of followons (HawkBoard, PandaBoard, etc.) Anyone familiar with the BeagleBoard will find the name quite descriptive.
Of course, the Pi has most of the mindshare, so the comparisons are inevitable even though the products really occupy two different niches - dirt cheap educational/hobbyist platform in the case of the Pi, relatively inexpensive open source embedded reference design/eval board in the case of the ZooBoards. And yes, if you've ever seen the prices for "real" embedded processor eval boards with OEM developer support, the ZooBoards are quite cheap.
...because it really looks like it is aimed at the mini/micro/nano-itx stuff. (which it is easily 25-30% cheaper than street price for a 3 year old board).
My two cents is this is putting nails in another coffin, VIA, and shoring the trenches to undercut the upcoming toys from AMD.
Being sales and marketing for the minnowboard must be one of the hardest jobs in the world. It has the hand of Jobs marked on it. (The triumph of marketing over specification).
Imagine the project is a drone for mapping.
5 volts, 2 Amps is a powerful start point, USB, HDMI, 10/100 Ethernet, 1GB memory, Sd to 64 Gb, mic and headset., WiFi. Out of the box 32 to 50 pounds sterling or 50 to 75$. On your desk with android OS. Millions under construction. A worthy successor to Pi. Powered by arm 9 and cortex 15 Quad core.
Imagine a project to provide engine management, transmission control, data acquisition and in car display systems
Most of the start platform requirement plus some more goodies exist on this board.
Now someone is going to suggest that the lower cost item is not as durable as its more expensive sibling. Please read up on IBMs excellent paper describing HACMP
Finally
PC on a stick is the smallest offering of this family and they are tiny.
They are so small that everyone's grandson can be flying an automated drone enabled helicopter with camera and standby RC handover for Xmas. All it requires is a smattering of coding experience.
Coding
Android SDK and emulator FOC from Google
Eclipse Editor FOC
Basic4Android if you are a Visual Studio Fan this is for you(A few $'s and worth it)
Now look at all of the supported devices with source code, then add to them the supported Pi projects and the porting from Pi to Android is elementary.
The barriers to the successful completion of the automation of tedious, repetitive tasks have now been removed, a project can be completed within the budget of the masses. Now I believe that this fits in with Bill Gates vision of computing
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
