Huawei kondigt Watch GT aan met accuduur van 2 weken
Huawei heeft tijdens een persevent in Londen een nieuw slim horloge aangekondigd dat in tegenstelling tot eerdere smartwatches niet onder Google’s Wear OS draait maar onder een eigen besturingssysteem genaamd RTOS. Gevolg is weliswaar het niet kunnen draaien van apps van derden maar daar staat een accuduur van 2 weken tegenover.
De Watch GT heeft een tijdloos ontwerp van glas, metaal en keramiek en is verkrijgbaar in twee uitvoeringen; een Sport met zwart plastic bandje en een Classic variant met bruin lederen polsband. Alhoewel beide uitvoeringen waterdicht (5ATM) zijn kan alleen met het sportieve Graphite Black-bandje gezwommen worden. Het scherm bestaat uit een rond AMOLED-paneel met diameter van 1,39 inch en een resolutie van 454 bij 454 pixels.
De Huawei GT Sport (zwart) en Classic (bruin leer)
Het ontbreken van Wear OS lijkt hem minder geschikt te maken als ‘slim horloge’. Notificaties van je telefoon kunnen alleen via Bluetooth doorgegeven worden omdat WiFi of 3G/4G ontbreekt. Hierdoor zijn wel flink langere batterijtijden mogelijk, volgens Huawei tot zelfs 2 weken. Schakel de hartslagmonitor uit en een maand moet haalbaar zijn. De hartslagsensor is trouwens wel het onderdeel waar Huawei met meest trots op is, zo lijkt. Deze scant continue je pols in tegenstelling tot iedere zoveel tijd. Wat de Huawei Watch GT precies met die data doet, waar het deze opslaat en hoe je deze kunt analyseren blijft onduidelijk uit Huawei’s persbericht. Ook de fitness-functies blijven ietwat onbelicht. Het horloge herkent wel diverse activiteiten waaronder hardlopen, zwemmen en fietsen maar onbekend blijft hoe en welke suggesties het geeft om je te verbeteren. Het redelijk eenvoudige karakter resulteert wel in een relatief goedkoop prijskaartje. De Sport-uitvoering met plastic bandje kost 199 euro, de Classic-variant met lederen bandje moet 249 euro kosten. Beiden liggen vanaf 23 oktober in de winkels.
Ontvang een e-mail wanneer de Watch GT verkrijgbaar is
Apple Launches New Beats Studio 3 Wireless ‘Skyline’ Collection
Apple this morning added a new “Skyline” collection of Beats Studio 3 Wireless Headphones to its online Apple Store, with a selection of new color options.
The Skyline Studio 3 Wireless Headphones come in Crystal Blue, Desert Sand, Midnight Black, and Shadow Grey. All of the colors feature gold accents with a gold beats logo and gold band.
Studio 3 Wireless Headphones feature Apple’s Pure Adaptive Noise Canceling technology, designed to block external noise when listening. They are also equipped with an Apple W1 chip that’s designed to make it easy to switch between devices when listening, and thanks to the efficiency of the chip, the Studio 3 Headphones last for 22 hours on one charge for all-day listening.
Like Apple’s other Studio 3 Wireless Headphones, the new Skyline collection is priced at $349.95. Skyline Beats are not yet available in Apple retail stores and must be ordered online, but should be arriving in retail locations soon.
Tag: Beats
Discuss this article in our forums
Review: Philips’ Hue Outdoor Lightstrip With HomeKit Integration Lights Up Your Backyard
Earlier this year, Philips debuted a new range of outdoor Hue lighting products that are designed to bring the iPhone-controlled multi-colored lights that you’re used to using inside to the outside.
There are outdoor spotlights, wall lanterns, pathway lights, outdoor bulbs, and for accent lighting, the Hue White and Color Ambiance Outdoor Lightstrip.
Measuring in at 80 inches, the $89 Outdoor Lightstrip is perfect for balconies, patios, backyards, and anywhere else you’d like to add gorgeous and functional accent lighting that’s also weatherproof.
Design
I’ve been using Hue products for years and I have a standard indoor Hue Lightstrip, and when I got the outdoor version, I assumed it’d be similar. It is, in the sense that it’s a long string of LEDs, but the similarities end there.
While the indoor Hue Lightstrip features LEDs that are uncovered, for weatherproofing reasons, the outdoor Hue Lightstrip’s LEDs are protected by an opaque silicone cover that diffuses the light and looks fantastic.
I’m a fan of my indoor Hue Lightstrips, but the way the outdoor version diffuses the light is superior, and I wish the indoor models looked like this. Because of the silicone-covered design, the Outdoor Lightstrip doesn’t need to be used as background accent lighting, it can be front and center on a pathway, along a railing, on a fence, or in other similar spots.
The Outdoor Lightstrips are available in 80 or 197 inch configurations. Unlike the indoor version, there is no adhesive involved here. Instead, the Outdoor Lightstrip is made from a flexible material that can be manipulated and shaped into the design that you need. It’s not going to fully bend, but it can be routed along a soft curve.
Hue’s Outdoor Lightstrip is fully weatherproof, so it’s going to hold up in rain, snow, and other harsh weather conditions. There are some limitations to be aware of if you live in an area with extreme temperatures, though, because it has an operational range of -4 degrees Fahrenheit to 113 degrees Fahrenheit.
It is a heavy duty setup with a large weatherproof power supply and an extra long two-piece cord, all of which is resistant to rain, dust, and other outdoor elements. I was tempted to use the Outdoor Lightstrip indoors because I loved the look of the diffuse light, but there is a lot of cord to deal with and no option to make it shorter.
The Outdoor Lightstrip measures in at .78 inches tall and .43 inches wide, so it’s slim enough to go just about anywhere. There is no adhesive involved because it likely wouldn’t stick outdoors. Instead, Philips supplies mounting brackets and screws that can be used to attach the Lightstrip to a fence, an overhang, or anywhere else.
I live in an apartment so my ability to permanently attach the Lightstrip to an outdoor area is limited, but I did install it on a set of shelves that I have to hold my plants, where it worked brilliantly. With the individual LEDs covered, the Outdoor Lightstrip looks like one continual light strand.
Though the light is diffuse, the Outdoor Lightstrip is bright enough to serve as atmospheric lighting for the outdoors. Total brightness will vary by color, with the lumen output rated at 900 lumens at 6500K (cool white light), 850 lumens at 4000K, 760 lumens at 2700K, and 740 lumens at 2000K (warm yellow light).
As with all White and Color Ambiance products, the Outdoor Lightstrip can be set at several different white light temperatures or to 16 million different colors in various shades and hues of red, orange, yellow, green, blue, and purple. Most of the colors are vibrant and accurate, though the purple shade is weak.
Some lightstrip options let you change the color of individual LEDs or sets of LEDs, but that’s not the case with Hue products. The lightstrip can only be set to one color at a time.
With the indoor Hue Lightstrips there are connectors at the end to add additional strands and the option to cut them in between LEDs for a more exact fit. The Outdoor Lightstrip does not support these features. You can’t cut it or connect another strand because doing so would negate the waterproofing.
Because there’s no option to size the Outdoor Lightstrip, it’s less flexible than the indoor version. You are limited to the two size options with nothing in between.
On the plus side, the weatherproofing works. As mentioned before, I set up the Outdoor Lightstrip on an outdoor shelf with plants, which were watered on a daily basis. The lightstrip, the cord, and the power adapter were repeatedly exposed to moisture for over a week with no problem.
Setup
To use individual Hue products like the Outdoor Lightstrip, you need an existing Hue setup with a Hue Bridge, which is used to control all of the lights. Hubs are available in Hue starter kits priced as low as $100 or on a standalone basis for $60.
If you already have a Hue Bridge, setting up the Outdoor Lightstrip is as simple as opening up the Hue app, selecting the Settings option, choosing “Light setup” and selecting “Add light.”
While Hue devices are all HomeKit-enabled, HomeKit control comes via the Bridge and Hue lights can’t be added through the Home app itself – you will need the Hue app.
If you don’t have a Hue Bridge set up, you’ll need to plug one in to your router, download the Hue app, and then follow the steps before you can use a Hue Lightstrip.
For the Outdoor Lightstrip, you’re going to need it to be within range of your Hue Bridge, which in turn needs to be connected to a Wi-Fi router, so it’s not going to be able to be located too far from the house. I tested mine on a patio outside, and while I have a mesh Wi-Fi system, the closest router was about 15-20 feet away and the signal was fine. The Outdoor Lightstrip can be up to 64 feet from the Hue Bridge or the next nearest light according to Philips’ documentation.
App and Functionality
The Outdoor Lightstrip works with other Hue lights in the Hue app, which is one way to control Hue lights, though there are many third-party app options as well.
The Hue app lets you organize your lights into rooms, name them, and control the brightness and the color. All of your installed Hue lights are listed on the Home screen, with quick options for changing brightness or turning them off all at once.
Tapping into a room provides a list of all of the lights, with additional quick access buttons for adjusting brightness or off/on state on a per light basis.
If you tap on one of the lights, like the Outdoor Lightstrip, there’s a color picker that lets you choose color and saturation for an array of different options at your fingertips. It will also display your most recently used color choices, and at the right, there are toggles to switch between the color palette and the white light palette.
Back at the main menu, there are options to set a selection of lights as a scene that can be recalled later, there’s a list of previously saved scenes, and there’s a palette for adjusting the colors of all your lights all at once.
There are some advanced options here, such as dragging the markers for multiple lights together to create lighting groups that can be adjusted all at once.
Along with the Home screen, the Hue app has a Routines section, used for creating automations for all of your Hue lights. You can use Routines to have the Outdoor Lightstrip come on or turn off at specific times, automatically at sunset/sunrise (with fading in and out), on a timer, or when you leave home or arrive home.
The app also features an Explore option for finding third-party Hue apps and products, and a Settings app for software updates, configuring accessories, changing rooms, renaming lights, and customizing widgets and the Apple Watch app.
With the Hue app, you can control your lights when you’re away from home so long as you’re signed in to your Hue account. Remote control is also available through HomeKit functionality, and the lights can also be set to come on automatically using several Hue or Home app methods.
Hue lights have been around for years now, which has given Philips a lot of time to refine the Hue app. It’s gone through growing pains, but as of now, the Hue app is functional, easy to use, and a useful way to control your Hue lights, including the Outdoor Lightstrip.
HomeKit
The Hue Bridge is HomeKit enabled, which means all Hue lights connected to it can be controlled like any other HomeKit device.
When adding a new Hue light to a Hue Bridge that’s connected to HomeKit, it’s available immediately as a HomeKit accessory. It can be accessed through the Home app, and there are options for changing the name of the lightstrip (which should be set to something easy to speak aloud to Siri), changing its room, or grouping it up with other accessories.
There are rudimentary controls for adjusting brightness and changing the color in the Home app that are fine to use in a pinch, but not as simple as the controls in the dedicated Hue app. There’s a slider for brightness and an adjustable color picker with the option to save your favorite colors for quick access.
With HomeKit’s Siri integration, you can ask Siri to turn the Outdoor Lightstrip on or off, dim the brightness, turn the brightness up, or set it to a specific color. Asking Siri to turn the Outdoor Lightstrip pink or blue, for example, works well to change the colors without needing to open up the Hue app.
Siri can set Hue devices to many standardized X11 color names, and it’s a useful way to get a specific color that you can’t necessarily get easily from the color picker within the Hue app.
Along with Siri controls, you can use HomeKit to create automations and control the Outdoor Lightstrip right alongside all of your other HomeKit products. So, for example, if you have a motion detecting HomeKit device, you can do things like set the lightstrip to come on automatically when motion is detected.
You can set up many of the same time-based automations through HomeKit that are available in the Routines section of the Hue app, but HomeKit offers more granular control on a light by light basis, while the Hue app does not.
Automations can be set based on geofencing options when people leave or arrive, a specific time of day, or based on another accessory.
Beyond HomeKit, the Outdoor Lightstrip works with Amazon Alexa, Google Assistant, Nest, and Samsung’s SmartThings.
Bottom Line
As a longtime owner of Hue lights, I love the look of the Outdoor Lightstrip, and I hope that the diffuse look created with the silicone cover is expanded to other Hue Lightstrip products in the future.
Outdoors, the Outdoor Lightstrip looks fantastic, and while it doesn’t put off enough light on its own to read by, it creates fantastic ambient lighting outdoors and can be used to light up pathways, balconies, patios, and more.
You can’t adjust the Outdoor Lightstrip’s size by cutting it or adding additional strips so it might be difficult to get a perfect fit, but this limitation allows for the waterproofing that keeps it safe outdoors.
If you have a Hue lighting setup already and have an outdoor space that’s close to the house and ideal for this kind of setup, the Outdoor Lightstrip isn’t going to be a disappointment.
If you’re new to Hue, you can’t quite start with the Outdoor Lightstrip, but the whole system is worth checking out if you’ve been looking for simple, attractive outdoor lighting.
As is typical with Hue lighting products, the Outdoor Lightstrips aren’t cheap, but they’re certainly well made and are able to hold up to continued exposure to outdoor conditions.
How to Buy
The Hue White and Color Ambiance Lightstrip Outdoor can be purchased from the Philips website for $89.99 for an 80 inch version. A 197-inch version is also available for $160.
Tag: Philips Hue
Discuss this article in our forums
App Store Bundles Gain Support for Mac Apps and Free Apps With Subscriptions
Apple today announced expanded features for App Store bundles, introducing support for both Mac apps and free apps with subscriptions for the first time.
Mac app developers are now able to create bundles of up to 10 apps, allowing customers to purchase multiple Mac apps at once at a discounted price.
Bundles like these have long been available through the iOS app store for purchasing multiple games or apps at one time, but until today, Mac App Store developers were not able to create similar bundles.
Apple is also now allowing app developers to set up bundles that include free apps with auto-renewable subscriptions, letting users purchase an app bundle and access multiple apps at one subscription price.
For free apps with subscriptions, each app in the bundle must have an approved auto-renewable subscription. If a user subscribes in one app, they must be able to access all other apps in the bundle at no additional cost. Learn about offering subscriptions across multiple apps.
Mac App Store developers have long wanted the ability to create app bundles, and the introduction of the feature follows the launch of the revamped macOS Mojave Mac App Store that makes it easier for customers to find and download new Mac apps.
Tag: App Store
Discuss this article in our forums
Apple Debuts Limited Edition Mickey Mouse Beats Solo 3 Wireless Headphones Created in Collaboration With Disney
To celebrate the 90th anniversary of Mickey Mouse, Apple has introduced special edition Mickey-themed Beats Solo 3 Wireless Headphones.
Priced at $329.95, the headphones are a gray color with a classic Mickey design. The headphones come with a matching gray felt carrying case, a collectible pin, and a decal sticker.
Celebrate “The True Original” with Mickey’s 90th Anniversary Edition Beats Solo3 Wireless headphones. Keep the magic going with award-winning sound and up to 40 hours of battery life. The cool-gray headphone features Mickey Mouse in the iconic pose seen on classic t-shirts that we all know and love. It also comes with a custom felt case inspired by the materials from Mickey Mouse ear hats, along with a collectible pin and decal sticker to honor the collaboration.
Mickey Mouse made his first appearance on November 18, 1928 to replace Oswald the Lucky Rabbit, one of Disney’s original characters. He made his debut in the short Plane Crazy, which was shown in short film Steamboat Willie, and from there, Mickey was in more than 130 films, both short and feature length.
Mickey Mouse is now one of the most iconic and well-known Disney characters, which is likely why Apple has designed Mickey-themed headphones. The Beats brand has done similar character collaborations in the past and once created Hello Kitty themed Solo Wireless Headphones and UrBeats.
The new Mickey Mouse headphones are listed as “Coming Soon” on Apple’s own website, but can be pre-ordered from Amazon. The official launch date on Amazon is listed as November 11, 2018.
Discuss this article in our forums
Izzbie review: a new kind of VPN
Data security is a hot topic. We’ve seen leaks from massive companies like Google and Facebook recently where people have had their data publically exposed. Many people are also concerned about third-party companies and the government snooping on their traffic. Because of these issues, Virtual Private Networks or VPNs have never been more popular.
The way most VPNs work is by creating a secure tunnel from your computer or phone to the VPN service’s servers. Your data travels along that tunnel and then once it gets to the servers, it heads to its intended location.
Read more: Should you use a VPN on your phone?
The reason for using a VPN numbers in the hundreds, if not thousands. First off is privacy. By using a secure connection and encrypting your data, it’s hard if not impossible for someone to snoop on what you’re doing. VPNs also allow you to get around region restrictions and ensure that your connection isn’t susceptible to a man-in-the-middle attack while connecting to a site where you use sensitive information, like your bank.
There are a number of VPN options out there. Many are paid like Private Internet Access. There are some free services out there, but as the saying goes, if you’re not the customer, you’re the product.
But, now there’s a new kid on the block. For the last couple of weeks, we’ve been testing out the new VPN service from Izzbie.
Izzbie is a different kind of VPN that leverages your home network to a create safe and convenient private network. There’s no third-party cloud or 2P2 software involved and it uses military-grade AES 256 encryption for your data.
Sounds amazing right? So, let’s take a look at how it works.
First up, you attach an Izzbie One box to your network. I have mine coming out of my switch. You don’t have to fiddle with putting it between your connection and modem or router, it just connects with an open ethernet port.
Then, you run through the setup process and log in. Walking through setup in the app can be a tad confusing, but Izzbie has informed us that the app will be getting a redesign to make it more functional and intuitive to use. The company includes setup instructions in the box and has a video walkthrough on its YouTube channel as well. We found that the setup, once properly executed, is done within minutes and pretty painless.
After your account is set up, its time to find the Izzbie box on your network. A simple tap of the “discover devices” button and you’re off. Once you are connected to the box, you’re done. Now, all of your data from your phone is flowing through your home internet connection.
At this point, you may be asking yourself why you’d even want to do this. I am by no means a networking guy and I had trouble wrapping my head around the advantages here. If I’m going to route all of my network data through my home internet, why wouldn’t I just stay connected to WiFi all the time?
The real magic of Izzbie is when you leave the house. Now, you’re not connected to your WiFi but your data is still running through your home network. And, you get the added benefit of encryption.
If you’re on a business trip in another country and want to watch your favorite Netflix shows during your downtime, this is perfect for getting around region restrictions. If you have a home NAS that can only be accessed from inside your network, boom, you’re good to go.
You can also expand your possibilities by adding another Izzbie box to your arsenal. During our testing, I had a friend set up a box at his house so I could test the functionality. It helped quite a bit as I was troubleshooting some Plex server issues I was experiencing. But, if your workplace requires a secure connection to remote in, this is also an excellent option.
Izzbie tells us that the real value of the system is hooking up multiple Izzbie One boxes. It allows users to connect multiple LAN networks into a virtual LAN. This allows users to get a LAN-like connection to all of the IT resources they need access to. It also allows user-to-network and network-to-network communications as if they’re all on the same network.
There are downsides here, but they minor. You are capped to the speed of the internet connection your box is connected to. Instead of pulling 80-100Mbps through my cellular connection, I was limited to 15 up and down because that’s my home internet speed’s upload limit. If you have fiber, you shouldn’t really run into this issue. Additionally, if you’ve ever used a standard VPN, you know how slow they can be.
Izzbie is generally much faster, especially when connecting to a NAS. You get a LAN-like connection, which is much much faster. This allows users to map the NAS as if they were inside the same LAN. The same benefits apply if you’re connecting to devices like IP cameras, printers, or servers.
And it is a little pricey. The almost $300 price tag does feel a little prosumer. But, you are getting all the hardware you need and a lifetime subscription. In a world where everything has a recurring cost to it, it’s nice to be able to buy something and just be done paying for it. And Izzbie is appropriate for a prosumer or small business environment, so those who are looking for a little bit more than your standard consumer-grade options will be getting your money’s worth.
Despite those minor annoyances, we really find a lot of value in just how simple the system can be to set up and use. You see a ton of value here when you compare Izzbie One to a VPN router. Izzbie doesn’t require router configuration, a static IP, or DDNS. Additionally, you never have to open any incoming ports to get Izzbie to work, a major plus for network security. All you need to is a single user login and you’re on your way.
If you’re maintaining a corporate network, there are a ton of advantages for you too. It’s super easy to add and remove users for the Izzbie network. The system also allows your users to connect to multiple locations with just one login, instead of juggling multiple logins. No more tickets to reset credentials!
We think this is a great start for Izzbie. There are some things that need addressed like the aesthetics of the app, but those changes are on the way.
We spoke to the CEO, Alan Hui, to get an idea of where the company and the service are going in the future.
He told us that Izzbie isn’t stopping here. It plans to bring IoT products to the market like smart door locks and cameras that all use the Izzbie platform. Since you’d be able to use these devices without requiring the cloud to relay the data transmission, it’ll be much faster and more secure to use. It will also allow you to connect to IoT products in multiple locations using the same login, something most cloud services can’t do at this time.
So, should you invest in Izzbie? If you’re looking for a different kind of VPN, we think so. There’s a lot of promise here and we’re really interested to see where the company goes. If you’d like to get in on the ground floor, you can pre-order by checking out Izzbie’s Indiegogo campaign.
Leaked benchmarks suggest rumored AMD GPU could be king of midrange graphics
A leaked benchmark result for a “generic” graphics card that seems likely to be the rumored RX 590 AMD has been working on has appeared on the 3DMark database. Although a brand-new card from AMD in over a year is exciting in its own right, the fact that it achieved a score of over 5,000 in Time Spy suggests it could be much more powerful than the GTX 1060 6GB, potentially making the RX 590 the best midrange GPU in the world.
It’s not available yet, or even officially announced, which does put a bit of a dampener on its plans for world domination. Still, the benchmark entry seems legitimate and backed up by earlier rumors of a supposed minor refresh of AMD’s RX 500-series cards, the evidence is mounting of a new AMD graphics card or cards showing up before Navi in 2019.
On top of the driver name outing this card as an RX 590, it seems more likely to be an enhanced version of the RX 580, rather than a new-generation with a designation like RX 680 or similar. It features the same 8GB of GDDR5 and the same memory bus clock of 2,000MHz. Where it does differ though, is on the core clock, which is 1,545MHz, compared to the RX 580’s typical clock of around 1,340MHz.
According to rumors, this clock speed bump is possible thanks to the new GPU being built upon a Polaris 30 design that uses a 12nm fabrication process, as per TechRadar. RX 500-series and Vega cards are built on a 14nm Polaris 20 process. The core clock increase is likely to be what’s most responsible for the new, higher score in 3DMark.
The new entry has a combined score of 5,028, with a graphics score of 4,759. A near-stock-clocked RX 580 paired with the same CPU achieved a combined score of 4,661, and a graphics score of 4,364. A GTX 1060 at near-stock-clocks and paired with the same CPU achieved a score of 4,248, with a graphics score of 3,954. There are better GTX 1060s that achieved higher clocks and scores, but even the most overclocked ones don’t reach the RX 590’s scores.
That’s the big takeaway from this news. That the RX 590 could become a true middle ground between the GTX 1060 and GTX 1070, something that Nvidia has yet to offer anything for. Although 1070 prices have come down a lot in recent months, with no comparably priced alternative for either in the RTX-series of cards, AMD’s RX 590 could plug an important hole in the GPU lineup — if it’s priced competitively, that is. But midrange products at competitive pricing is AMD’s bread and butter in the graphics market.
Look out for an update to our best graphics cards roundup if and when this card debuts.
Editors’ Recommendations
- AMD RX 500 graphics card refresh may launch in 2019 with die shrink, clock boost
- Nvidia vs. AMD
- The best graphics cards
- Email shows Nvidia’s GTX 11 Series will start being released in late August
- 3DMark’s new Night Raid benchmark gives integrated graphics their own test
What is deep learning?
Deep learning is a particular subset of machine learning (the mechanics of artificial intelligence). While this branch of programming can become very complex, it started with a very simple question: “If we want a computer system to act intelligently, why don’t we model it after the human brain?”
That one thought spawned many efforts in past decades to create algorithms that mimicked the way the human brain worked—and that could solve problems the way that humans did. Those efforts have yielded valuable, increasingly competent analysis tools that are used in many different fields.
The neural network and how it’s used
via Wikipedia
Deep learning gets its name from how it’s used to analyze “unstructured” data, or data that hasn’t been previously labeled by another source and may need definition. That requires careful analysis of what the data is, and repeated tests of that data to end up with a final, usable conclusion. Computers are not traditionally good at analyzing unstructured data like this.
Think about it in terms of writing: If you had ten people write the same word, that word would look very different from each person, from sloppy to neat, and from cursive to print. The human brain has no problem understanding that it’s all the same word, because it knows how words, writing, paper, ink, and personal quirks all work. A normal computer system, however, would have no way of knowing that those words are the same, because they all look so different.
That brings us to via neural networks, the algorithms specifically created to mimic the way that the neurons in the brain interact. Neural networks attempt to parse data the way that a mind can: Their goal is to deal with messy data—like writing—and draw useful conclusions, like the words that writing is attempting to show. It’s easiest to understand neural networks if we break them into three important parts:
The input layer: At the input layer, the neural network absorbs all the unclassified data that it is given. This means breaking down the information into numbers and turning them into bits of yes-or-no data, or “neurons”. If you wanted to teach a neural network to recognize words, then the input layer would be mathematically defining the shape of each letter, breaking it down into digital language so the network can start working. The input layer can be pretty simple or incredibly complex, depending on how easy it is to represent something mathematically.
via Wikimedia
The hidden layers: At the center of the neural network are hidden layers—anywhere from one to many. These layers are made of their own digital neurons, which are designed to activate or not activate based on the layer of neurons that precedes them. A single neuron is a basic “if this, then that“ model, but layers are made of long chains of neurons, and many different layers can influence each other, creating very complex results. The goal is to allow the neural network to recognize many different features and combine them into a single realization, like a child learning to recognize each letter and then forming them together to recognize a full word, even if that word is written a little sloppy.
The hidden layers are also where a lot of deep learning training goes on. For example, if the algorithm failed to accurately recognize a word, programmers send back, “Sorry, that’s not correct,” and the algorithm would adjust how it weighed data until it found the right answers. Repeating this process (programmers may also adjust weights manually) allows the neural network to build up robust hidden layers that are adept at seeking out the right answers through a lot of trial and error plus, some outside instruction — again, much like how the human brain works. As the above image shows, hidden layers can become very complex!
The output layer: The output layer has relatively few “neurons” because it’s where the final decisions are made. Here the neural network applies the final analysis, settles on definitions for the data, and draws the programmed conclusions based on those definitions. For example, “Enough of the data lines up to say that this word is lake, not lane.” Ultimately all data that passes through the network is narrowed down to specific neurons in the output layer. Since this is where the goals are realized, it’s often one of the first parts of the network created.
Applications
If you use modern technology, chances are good that deep learning algorithms are at work all around you, every day. How do you think Alexa or Google Assistant understand your voice commands? They use neural networks that have been built to understand speech. How does Google know what you’re searching for before you’re done typing? More deep learning at work. How does your security cam ignore pets but recognize human movement? Deeping learning once again.
Anytime that software recognizes human inputs, from facial recognition to voice assistants, deep learning is probably at work somewhere underneath. However, the field also has many other useful applications. Medicine is a particularly promising field, where advanced deep learning is used to analyze DNA for flaws or molecular compounds for potential health benefits. On a more physical front, deep learning is used in a growing number of machines and vehicles to predict when equipment needs maintenance before something goes seriously wrong.
The future of deep learning
The future of deep learning is particularly bright! The great thing about a neural network is that it excels at dealing with a vast amount of disparate data (think of everything our brains have to deal with, all the time). That’s especially relevant in our era of advanced smart sensors, which can gather an incredible amount of information. Traditional computer solutions are beginning to struggle with sorting, labeling and drawing conclusions from so much data.
Deep learning, on the other hand, can deal with the digital mountains of data we are gathering. In fact, the larger the amount of data, the more efficient deep learning becomes compared to other methods of analysis. This is why organizations like Google invest so much in deep learning algorithms, and why they are likely to become more common in the future.
And, of course, the robots. Let’s never forget about the robots.
Editors’ Recommendations
- Machine learning? Neural networks? Here’s your guide to the many flavors of A.I.
- What is an artificial neural network? Here’s everything you need to know
- What the heck is machine learning, and why is it everywhere these days?
- MIT’s latest A.I. is freakishly good at determining what’s going on in videos
- How A.I. can defeat malware that doesn’t even exist yet
What is deep learning?
Deep learning is a particular subset of machine learning (the mechanics of artificial intelligence). While this branch of programming can become very complex, it started with a very simple question: “If we want a computer system to act intelligently, why don’t we model it after the human brain?”
That one thought spawned many efforts in past decades to create algorithms that mimicked the way the human brain worked—and that could solve problems the way that humans did. Those efforts have yielded valuable, increasingly competent analysis tools that are used in many different fields.
The neural network and how it’s used
via Wikipedia
Deep learning gets its name from how it’s used to analyze “unstructured” data, or data that hasn’t been previously labeled by another source and may need definition. That requires careful analysis of what the data is, and repeated tests of that data to end up with a final, usable conclusion. Computers are not traditionally good at analyzing unstructured data like this.
Think about it in terms of writing: If you had ten people write the same word, that word would look very different from each person, from sloppy to neat, and from cursive to print. The human brain has no problem understanding that it’s all the same word, because it knows how words, writing, paper, ink, and personal quirks all work. A normal computer system, however, would have no way of knowing that those words are the same, because they all look so different.
That brings us to via neural networks, the algorithms specifically created to mimic the way that the neurons in the brain interact. Neural networks attempt to parse data the way that a mind can: Their goal is to deal with messy data—like writing—and draw useful conclusions, like the words that writing is attempting to show. It’s easiest to understand neural networks if we break them into three important parts:
The input layer: At the input layer, the neural network absorbs all the unclassified data that it is given. This means breaking down the information into numbers and turning them into bits of yes-or-no data, or “neurons”. If you wanted to teach a neural network to recognize words, then the input layer would be mathematically defining the shape of each letter, breaking it down into digital language so the network can start working. The input layer can be pretty simple or incredibly complex, depending on how easy it is to represent something mathematically.
via Wikimedia
The hidden layers: At the center of the neural network are hidden layers—anywhere from one to many. These layers are made of their own digital neurons, which are designed to activate or not activate based on the layer of neurons that precedes them. A single neuron is a basic “if this, then that“ model, but layers are made of long chains of neurons, and many different layers can influence each other, creating very complex results. The goal is to allow the neural network to recognize many different features and combine them into a single realization, like a child learning to recognize each letter and then forming them together to recognize a full word, even if that word is written a little sloppy.
The hidden layers are also where a lot of deep learning training goes on. For example, if the algorithm failed to accurately recognize a word, programmers send back, “Sorry, that’s not correct,” and the algorithm would adjust how it weighed data until it found the right answers. Repeating this process (programmers may also adjust weights manually) allows the neural network to build up robust hidden layers that are adept at seeking out the right answers through a lot of trial and error plus, some outside instruction — again, much like how the human brain works. As the above image shows, hidden layers can become very complex!
The output layer: The output layer has relatively few “neurons” because it’s where the final decisions are made. Here the neural network applies the final analysis, settles on definitions for the data, and draws the programmed conclusions based on those definitions. For example, “Enough of the data lines up to say that this word is lake, not lane.” Ultimately all data that passes through the network is narrowed down to specific neurons in the output layer. Since this is where the goals are realized, it’s often one of the first parts of the network created.
Applications
If you use modern technology, chances are good that deep learning algorithms are at work all around you, every day. How do you think Alexa or Google Assistant understand your voice commands? They use neural networks that have been built to understand speech. How does Google know what you’re searching for before you’re done typing? More deep learning at work. How does your security cam ignore pets but recognize human movement? Deeping learning once again.
Anytime that software recognizes human inputs, from facial recognition to voice assistants, deep learning is probably at work somewhere underneath. However, the field also has many other useful applications. Medicine is a particularly promising field, where advanced deep learning is used to analyze DNA for flaws or molecular compounds for potential health benefits. On a more physical front, deep learning is used in a growing number of machines and vehicles to predict when equipment needs maintenance before something goes seriously wrong.
The future of deep learning
The future of deep learning is particularly bright! The great thing about a neural network is that it excels at dealing with a vast amount of disparate data (think of everything our brains have to deal with, all the time). That’s especially relevant in our era of advanced smart sensors, which can gather an incredible amount of information. Traditional computer solutions are beginning to struggle with sorting, labeling and drawing conclusions from so much data.
Deep learning, on the other hand, can deal with the digital mountains of data we are gathering. In fact, the larger the amount of data, the more efficient deep learning becomes compared to other methods of analysis. This is why organizations like Google invest so much in deep learning algorithms, and why they are likely to become more common in the future.
And, of course, the robots. Let’s never forget about the robots.
Editors’ Recommendations
- Machine learning? Neural networks? Here’s your guide to the many flavors of A.I.
- What is an artificial neural network? Here’s everything you need to know
- What the heck is machine learning, and why is it everywhere these days?
- MIT’s latest A.I. is freakishly good at determining what’s going on in videos
- How A.I. can defeat malware that doesn’t even exist yet
Qualcomm’s ‘Snapdragon 1000’ could bring octa-cores to Windows laptops
New details have emerged about the rumored, super-powerful Qualcomm Snapdragon 1000 CPU that is designed to take on Intel chips in the Windows laptop space. Although no official announcement of such a chip has been made, a new report suggests that it will bring the octa-core design most commonly found in smartphones and tablets, to Windows laptops, offering four high-powered cores for performance, and four lower-powered cores for efficiency.
Over the past few years, Qualcomm has gone from developing smartphone-exclusive processors with on-board graphics, to creating chips that are far more capable. Its Snapdragon 835 and 845 made their way into both high-end smartphones and tablets, and joined the Snapdragon 850 platform in the “Always Connected” Windows laptops, too. The Snapdragon 1000, however, would be far more capable than even those and could make such Always Connected devices far more competitive in the larger laptop market.
According to the report from WinFuture, translated by BGR, the Snapdragon 1000 is an eight-core CPU based on the Cortex-A55 and Cortex A76 architecture. Four of the cores would be clocked at 3.0GHz or higher, while the other four would sit at 1.8GHz and would be used to extend battery life when the system is idling or not handling performance-intensive tasks. The chip is also said to support LPDDR4X memory, which could have a dramatic improvement in memory bandwidth.
On top of the new, high-powered CPU cores, a brand-new GPU, known as the Adreno 680, is being added to the Snapdragon 1000. It should have a higher clock speed than the GPU cores used in the Snapdragon 835 and 845 chips. These improvements lead to a much denser, transistor-packed chip. The new report suggests that the Snapdragon 1000 will pack as many as 8.5 billion transistors under the hood.
Despite all of this higher-powered hardware, the thermal design power of the Snapdragon 1000 is said to only be 15w, which makes it competitive with some of Intel’s ultra-low wattage seventh-generation CPUs.
No news has yet surfaced on which laptop manufacturers may be the first to adopt the new Qualcomm chips, but BGR suggests that Asus is rumored to have a new laptop design in the works with them in mind called Primus.
If you don’t fancy waiting for that, here’s our list of the best laptops available right now, and a look at some of the best budget laptops, for all of your low-power needs.
Editors’ Recommendations
- Here’s everything you need to know about Intel’s 9th-gen chips
- Intel’s 9th-generation processor could launch next month with 8 cores
- Qualcomm’s Snapdragon 850 chip appears in benchmarks with improved performance
- Acer brings Intel’s 8-core, 9th-gen processor to its Predator Orion desktops
- Apple’s new MacBook Air with Intel’s 8th-gen processors expected this year
