If you’re playing Hackaday Buzzword Bingo, today is your lucky day! Because not only does this article contain “Pi 3” and “IoT”, but we’re just about to type “ESP8266” and “home automation”. Check to see if you haven’t filled a row or something…
Seriously, though. If you’re running a home device network, and like us you’re running it totally insecurely, you might want to firewall that stuff off from the greater Interwebs at least, and probably any computers that you care about as well. The simplest way to do so is to keep your devices on their own WiFi …read more
Movidius chips have been showing up in quite a few products recently. It’s the company that helps DJI’s latest drone avoid obstacles, and FLIR’s new thermal camera automatically spot people trapped in a fire, all through deep learning via neural networks. It also signed a deal with Google to integrate its chips into as-yet-unannounced products. Now, the chip designer has a product it says will bring the capacity for powerful deep learning to everyone: a USB accessory called the Fathom Neural Compute Stick.
The Fathom contains the Myriad 2 MA2450 VPU paired with 512MB of LPDDR3 RAM. The Myriad 2 is the chip found in the previously mentioned DJI and FLIR products. It’s able to handle many processes simultaneously, which is exactly what neural networks call for. Because it’s specifically designed for this — its architecture is very different from the GPUs and CPUs that typically handle processing — it offers a lot of grunt without requiring much power. It can handle up to 150 gigaFLOPS (150 billion floating-operations per second) while consuming no more than 1.2 watts.
Unlike Tegra’s solutions for deep learning, the Fathom isn’t a standalone system. The idea is you plug it into the USB 3.0 port of any system running Linux to get a "20-30x performance improvement in neural compute." You can use the Fathom to rapidly prototype neural networks, moving to something with a lot more power once you’re ready to deploy.
Of course, this is neural networking, so it’s not that simple. The Fathom accepts networks defined in Caffe and TensorFlow (two frameworks popular in deep learning circles) and their accompanying datasets. You need to use a Movidius tool to execute the network on the Myriad 2 chip, where it’ll run natively while sipping power. At first glance, it’s a very similar process to CUDA and cuDNN (Nvidia’s system for handing off neural networks to its graphics cards). That said, the whole point of Fathom is it can be used in an environment where you don’t have expensive graphics cards and processors.
The Fathom is a very interesting device. As anyone that’s attempted to run even a basic neural network on an underpowered machine will tell you, it’s slow going. At present, the best way to prototype a network is using a cloud-based system, tapping into computing power far away. Being able to add a decent amount of compute to a regular laptop could simplify and reduce the cost of building a network massively.
But the potential for Fathom doesn’t end there. It could prove very useful for robotics, drones, and the maker community at large. With a Fathom connected to a Raspberry Pi, for example, you could easily add some very advanced computer vision capabilities to something like a GoPro. The long game, of course, is to persuade more manufacturers to add Myriad chips into their devices, but something like the Fathom is a key step along the way.
DJI’s obstacle avoidance is powered by the same chip as the Fathom.
The AI community has reacted positively to the announcement. Facebook’s Director of Artificial Intelligence Dr. Yann LeCunn said he’s "been hoping for a long time that something like Fathom would become available … With Fathom, every robot, big and small, can now have state-of-the-art vision capabilities." while Google’s AI Technical Lead Pete Warden said that "Fathom goes a long way towards helping tune and run these complex neural networks inside devices."
While some organizations are being receiving their Fathoms now, the Neural Compute Stick won’t go on general sale until this winter. There’s no firm price yet, but we’re told it’ll be less than $100.
HPC blog The Asian Student Supercomputer Competition (ASC) is the closest thing to true stock car racing that you’ll find on the Student Cluster Competition international circuit.…
We’ve been hearing for years about all the ways the “internet of things” is supposed to revolutionize the world. When is that supposed to happen, though? Sure, you can get a few smart-things […]
Tile’s Bluetooth-connected trackers are made to keep you from misplacing all kinds of stuff — keys, bags, phones, laptops, whatever. Now Land Rover is building the tech into a vehicle that makes it pretty much impossible to leave your important items behind. In the 2017 Discovery Sport Tile will be a part of its InControl apps setup, ready to alert the driver if tagged items aren’t in the car. Maybe you’ve never shown up to the airport one bag short, but for the rest of us, it’s one way to avoid an awkward situation.
It’s also able to find items that are inside the car, by locating them and activating a 90db alert sound. The 2017 edition of the SUV has the option of a 10.2-inch touchscreen with a 21:9 aspect ratio (and no hard buttons whatsoever) to control the apps including Tile, and a higher-resolution screen allowing dual-screen use by driver and passenger without reducing the quality.
Windows/Mac/Linux/Android: One of Microsoft’s Office 365 program chief advantages over open source alternatives is the ability to sync documents via the cloud so you can edit them everywhere. Open365 has stepped up to finally match this feature set.
The content below is taken from the original (What’s the Future of the IT Pro?), to continue reading please visit the site. Remember to respect the Author & Copyright.
If you need help jogging your memory, you might try your hand at drawing. A recent study found that we remember items better when we draw them rather than write them down.
It’s hard to swing a stick in Silicon Valley without hitting startups created by ex-Google employees determined to bring their clever (though only sometimes successful) ideas to light. That’s good for innovation, but lousy for Google — and the search giant now appears bent on doing what it can to keep those curious minds in-house. The Information‘s sources understand that Google is creating Area 120, a startup incubator that would let some employees pursue their "20% projects" (those personal projects Google allows in a fifth of your working hours) full-time. Anyone wanting to sign up would submit a business plan and, if accepted, spend several months working solely on that idea. You could scratch that inventor’s itch without worrying that you’ll lose your cushy Google job if it it doesn’t pan out.
The company isn’t confirming anything yet, and it’s not clear exactly when Area 120 would be open for business. Having said that, an incubator wouldn’t be surprising. Google’s existing attempts at fostering a startup-like culture don’t always stop people from jumping ship — just ask the ATAP team, which just lost its leader to Facebook. And in a few cases, defectors go on to found companies that directly undermine Google. Kevin Systrom, for instance, worked two years at Google before the eventual creation of Instagram and its acquisition by (you guessed it) Facebook.
The catch, as The Information notes, is that this idea nursery might not be enough. It’s relatively easy to raise funds in the current Silicon Valley climate, and the freedom of going it alone may prove more tempting than Google’s safety net. However, it could be worthwhile for the internet pioneer if even one big new idea stays within its walls.
We all love to upload our images to social media. Be it to show our professional work, get critique from our peers or just to show our friends and family our latest shots, between us, we upload tens of millions of images per day to social media sites. As photographers, of course, we strive to […]
Tape still not dead: it will die in the year N where N is this year plus 1
Google and Iron Mountain are trying to hasten the never-quite-imminent death of tape as a storage medium with an LTO-to-cloud migration collaboration.…
Acer hit the stage a small event in the shadow of Manhattan’s One World Trade to show off some notebooks, notebooks and also some notebooks. Oh, and one more thing: notebooks, notebooks, and notebooks (also there were some notebooks). The star notebook amongst the notebooky notebooks is the company’s latest take on the Chromebook — the straightforwardly-named Chromebook 14… Read More
People who live in countries with a strict nationwide internet filter always come up with ways to get around it. In Iran, according to Wired, people are using satellite TV and a free anti-censorship system called Toosheh. While Iranians do use VPN to bypass the filter, their crippling internet speeds make it hard to stream videos or download bigger files. The system gives them a way to get 1GB of data within 60 minutes. Users simply have to plug a USB stick into the set-top box, access Toosheh’s channel that doesn’t show anything besides text instructions and set the receiver to record.
Once the download’s done, they plug the USB stick into their computers and use a tool to decode and unpack its contents. The Net Freedom Pioneers team (Tooshe’s creators) includes various files in their daily package, including TED talks, YouTube/music videos and news clips from banned websites, among others. Wired says the team makes it a point to include a mix of "entertainment, education and human-rights focused material" in each package. Obviously, it’s a poor alternative to having access to everything on the internet — when was the last time you’ve had to work that hard for YouTube video? — but it sure beats never being able to take a peek behind the censorship filter.
Slowly but surely, French startup Le Wagon is becoming a leading coding bootcamp in Europe. Not only the company is executing quickly and expanding to a bunch of new cities, but Le Wagon is also refining its approach to coding bootcamps by using technology as much as possible. Le Wagon started in Paris more than two years ago. Only 15 students took part in the first class. I covered the… Read More
It seems nothing is safe from the Internet of Things revolution. The latest attempt to make a household object smarter is Peggy, a connected clothes peg by Australian detergent brand OMO. A few basic components sit inside its orange shell, including a thermometer, UV sensor and humidity detector. These track the weather outside and, in the case of impending rain, trigger timely alerts to your phone over WiFi. So if the clouds roll in and you’re busy with something else — maybe you’re out of the house, but someone else is indoors — you won’t be left with soggy washing.
It’s a bit of a PR stunt. Most of this information could be obtained with a decent weather app or home weather station, after all. To sell the idea, OMO has developed some scheduling tools for the companion app which tell you the best time to put on a wash. Plug in the time it usually takes for your washing machine to complete a load, and it’ll explain how many hours (or minutes, if you’re unlucky) you’ll have afterwards before the next downpour. It’ll even send a notification when your washing machine is done, just in case you need an extra reminder to take out your digs.
OMO’s Peggy is currently in testing — if you’re curious and live in Australia, you can submit your details to ensure you’re “first in line” when it’s actually available. Otherwise, you can do what homeowners have been doing for, well, centuries, and keep one eye on the sky when your washing is hanging outside. Inefficient maybe, but it does save managing yet another device that needs to be charged up and connected to the internet.
Capital One is a huge organization with lots of compliance issues related to being a financial services company. It also happens to be an Amazon Web Services customer and it needed a tool to set rules and policies in an efficient way around AWS usage. Last July it started developing the tool that would become Cloud Custodian; today it announced at an AWS event in Chicago that it was making… Read More
Mesosphere‘s Data Center Operating System (DC/OS) aims to allow developers and admins to treat a data center as a single computer that runs applications in software containers. It’s based on a number of open-source projects, including the Apache Mesos cluster manager and projects like the Chronos scheduler and the Marathon container orchestration platform. Now, Mesosphere is… Read More
SODAQ’s “LoRaONE” is a tiny Arduino-compatible board with an Atmel SAMD21 MCU, plus a LoRa radio and a GPS, sensors, micro-USB, and optional Grove add-on. Dutch embedded firm SODAQ (SOlar powered Data AcQuisition) has almost doubled its $22,740 Kickstarter funding goal for its Arduino compatible LoRaONE SBC, aimed at low-power LoRa wireless applications. The cheapest […]
What do you do if even the smallest Arduino boards (or their clones) are too big for your homebrew project? If you’re Johan Kanflo, you find a way to make them even smaller. His AAduino project turns the already miniscule Tiny328 Arduino clone into an even smaller computing device that’s about as big as an AA battery. Through creative wiring, it even fits inside a typical battery holder and draws power from the batteries in the remaining slots. He had to underclock the processor to extend to the battery life, but it’s otherwise as capable as its normal counterparts.
You aren’t about to buy a ready-made AAduino, but you don’t have to. Kanflo has posted instructions and schematics both on his own site and on GitHub, so you can replicate his invention yourself. This is mainly useful if you’re building an extra-compact gadget (Kanflo needed this for a radio node, for instance), but it shows that even daunting size requirements can be solved with a little ingenuity.
I can trace my computing history along a path of display milestones. The 10-inch VGA monitor that came with my Packard Bell desktop in the ’90s was a huge leap forward from the low-res Apple II displays I used in school. Then there was the 20-inch Sony Trinitron flat CRT that I brought to college with my first custom-built desktop, which served as my dorm entertainment center for years. These days, I rock two 24-inch 1080p LCDs at home, and I’m eyeing an ultra-wide screen upgrade at some point. After spending several hours with Virtual Desktop, a $15 app developed by Guy Godin for the Oculus Rift and HTC Vive that brings the full Windows experience to VR, it seems like we’re close to yet another display revolution.
Imagine having your entire Windows desktop projected on a movie theater screen while sitting in the best seat in the house, and you’ll have some idea of what Virtual Desktop offers. Instead of a darkened theater, though, the display floats atop your choice of backgrounds (I really enjoy the detailed space environments). It brings to mind fantastical interfaces from science fiction — in particular, the trippy augmented reality setup from the game Heavy Rain.
There’s a definite "wow" factor to Virtual Desktop. After installing and running the app, I just slipped on my Rift headset and was presented with an impressive recreation of my Windows setup. Once you’re in virtual view, you can resize the size and distance of the desktop, from the equivalent of sitting six feet away from a 60-inch display, to having it tower before you as if it were on a enormous IMAX screen. I found it most comfortable to have the virtual screen completely fill up my field of vision, even though that also involved occasionally moving my head around to focus on specific things.
SteamVR’s "desktop theater mode" on the HTC Vive also lets you use Windows in VR, but in my testing it was hard to actually read text and use Windows as you normally would. It’s mainly meant for playing non-VR games and movies in your headset. In comparison, I had no trouble going through my email, browsing the web and reading long articles with Virtual Desktop. It was so comfortable, in fact, that I was able to use it for hours on end.
Virtual Desktop is particularly impressive when it comes to watching videos, since you can freely make the screen larger for a more cinematic experience. Of course, you won’t get the full resolution of 1080p HD videos, since both the Rift and Vive are limited by their displays, which only offer a resolution of 1,280 by 1,080 pixels per eye. Videos still looked sharp and clear in full screen mode, though. And after being underwhelmed by the virtual theater apps on both the Rift and Vive, which seem more focused on reminding you that you’re in a VR environment than actually making videos look good, I was surprised to find I actually enjoyed watching things with Virtual Desktop.
There’s also support for viewing 360-degree videos and photos with the app, which is useful for quickly jumping into an immersive clip from YouTube. This is one area that the app needs to improve on, though. To plug in a video, you need to copy a URL to your clipboard and insert that into Virtual Desktop’s settings app. It’ll then proceed to download the video and play it instantly. You can also open up 360-degree videos already on your computer, but it’s far more likely you’ll be grabbing content from YouTube and other sites.
The big problem with the app? You still have to rely on your keyboard and mouse to navigate Windows, which will definitely be an issue if you don’t touch-type well. Remember, you can’t easily see what’s going on in front of you. Virtual Desktop shows why something like the Vive’s webcam, which lets you quickly view the outside world, is essential for all VR headsets. I did most of my testing with the Oculus Rift, so I was forced to either type very carefully, or remove the headset if I needed to hit a specific key. (Function keys, in particular, are hard to touch-type.)
I was able to write several lengthy emails while wearing the Rift, but it always took a few seconds for my fingers to readjust to the key positions after taking my hands away. That ended up being a bigger issue for entering web addresses or numbers of any kind, since those involve hitting keys that aren’t in my normal touch typing range. (My slim wireless keyboard doesn’t have a number pad, though that probably would make my life easier.) Mousing around Windows in Virtual Desktop was fine, and the app also made something mundane like photo editing feel like a revelatory experience. Cropping an recoloring a giant image felt akin to working on a giant Jackson Pollock-esque canvas.
I won’t be giving up my dual monitor setup anytime soon, but I was surprised by how well my workflow transferred to Virtual Desktop. I could see creative types preferring it to a restrictive monitor setup, especially for things like video and audio editing, where you normally have dozens of tracks to juggle at one. Just like the first batch of VR hardware, the app feels like a sign of things to come. It won’t be too long until we can just slip on a pair of glasses to have a full desktop environment emerge from our smartphones.
