Intel Readies for Internet of Things Invasion with Linux | Hacker News Reader

Intel Readies for Internet of Things Invasion with Linux

linux.com

39 pointshepha197912 comments

I was pretty excited about the Galileo[1] board when it was announced (and a bit disappointed they laid me off before I could buy some at a discount). Might still buy one retail, since it's a good match for a project I have. It is not trying to be an x86 Raspberry Pi, as it has no video and more I/O, though I consider it a mistake to hang most of the GPIO pins off SPI. I also think they ought to have ensured before release that a wider range of OSes ran on it — various RTOSes, free and otherwise, and FreeDOS, to make it a drop-in replacement for the ol' PC-with-parallel-port.

[1] http://arduino.cc/en/ArduinoCertified/IntelGalileo

Intel really missed the boat on the cell phone front. I doubt they are excited about missing out in the IOT market as well. Look for more announcements from them in this area in the weeks/months to come.

It's funny how Intel and Microsoft were the ones calling for the Internet of Things all the time, and now that we are finally building it, they have no product fit for it.

umm... Intel DOES have a product fit for it, no? I mean isn't that what the post is about?

I must admit I'm very confused by the "Internet of Things". It seems largely to be a ploy by big data and cloud infrastructure types to hype investment in their sectors to absorb information from dumb yet internet connected sensors, whereas what it should be is smart end devices, yet conspicuously there isn't much in the way of development in the edge devices, while the cloud services are all ready and waiting, to do . . . what exactly?

With the NSA stuff having tarnished the word "cloud" maybe the marketers are hoping Internet of Things will divert attention away from the fact that they're suggesting even more invasion of privacy.

It is a nebulous concept but I believe it is a combination of smart end user devices and data gathering sensors... obviously in the IT world any new (relatively speaking) concept gets hijack for marketing by vendors.

Cheap low power chips that could be embedded in everything from your clothes to your pills I believe is what Intel is aiming at.

Yeah. Of course, what they're actually achieving is - predictably - rather less low power than existing embedded ARM solutions. If you want a good laugh take a look at the spec page for Quark sometime.

Don't forget about the "Quantified Self" marketing push, and of course every device syncs only to your mobile device via bluetooth. No PC required (or supported), because then you might not let us harvest all your data invisibly through our app!

You are as confused as I (still am) about the cloud :-). It's a cheap, stupid marketing ploy that honestly doesn't mean anything. Originally, it may have referred to IP(v6)-enabled small home automation devices, like smart light dimmers, distributed sensor-driven HVAC systems and the like, with obvious advantages. Half of the things marketed under the IoT mantra are too expensive and have way too much computational power to be of any use in this context.

From my view, it's the chipmakers looking at the high-end of the market getting locked up by Apple, Samsung, and the like.

These chipmakers need the growth, but the low-end chips are too cheap to make the margins add anything substantial. So, hey!, let's see if we can get the low-end chips into EVERYTHING and sort it out later.

There are obviously industrial and commercial applications aplenty for this, but does every lightbulb in your house need to be networked?

Earlier this year I did some development for a hardware designer who has been getting into energy-harvesting. He had a Bluetooth-LE profile miroprocessor running wired up to an energy-harvesting chip, meaning that we were able to build a network of Bluetooth-LE devices that didn't require any power supplies.

These prototypes are working (the app is made to render the whole kaboose as a point-cloud) and as far as I can tell, the only reason these things haven't hit the market yet, is timing. He's also not the only designer working on such products - as a result of contact with the LE/Infinergy industry, I see quite a few other startups attempting to 'capture' this space before it launches.

So I predict it won't be long until we have such devices out there. I'm already using one of the prototypes to find my keys .. only problem is, I can't use it to find my phone, where the app is installed .. ;)

Remember that networked hard-drive by sea gate? We could probably use that here too. Instead of storing all data in cloud (and thus increase network delay) we could simply store it in a local hard drive which is accessed by these devices directly via LAN.

Though i am confused about why the focus just sensors? I see it as micro-controllers connected to the internet, thus enabling a lot more things.

Reply on news.ycombinator.com