google Archives - COBRA softwares
20 Oct

13 Google Maps for Android Tricks That’ll Change How You Navigate

Are you really using all that Google Maps has to offer? Now it’s time you did! Google Maps has grown over the years and so have its features. It is now way more than just a digital replica of a paper map and you can get way more out of. But most of us are only scratching the surface of what they have to offer. There are some tricks hiding that you never ran into that will change the way you navigate.

Sure, you can use it like a map, but its way more powerful if you learn how to save places offline, taking advantage of navigation, improve the information by adding details, and mastering other tricks, etc. Below are 13 tips and tricks for getting most out of the Google Maps on your Android. Some of these are big time savers, while others simply give you more options in configuring Maps to work the way you want it.

1. Long Press the blue button to Navigate Quicker

When you look up for a location in Google Maps, a blue button appears in the bottom right that you can tap to view the different ways of navigating there.

Sometimes you don’t really care how you get there, but that you get there as efficiently as possible. Hence, to quickly launch into navigation, simply long-press the blue button when you select a location — Maps will then start navigation to that place from your current location and take the fastest route possible. No messing with start points or routes, just long-press that button and go. It works for all navigation options like public transit and biking as well.

2. Double Tap and Hold to Zoom

This is a simple one, but it’s a game-changer for one-handed usage. Google Maps doesn’t require you pinch the screen with two fingers in order to zoom. Instead, try double tapping on a location to zoom partially in on that spot.

In addition, you can hold that second tap and slide your finger up to zoom in or slide your finger down to zoom out. So that’s “tap, tap, and hold, swipe up or down.”

3. Add Labels to Locations

You probably aware that you can “Star” items to save them for later in My Places, but do you also know that you can add labels to places? Yes, you can create labels for different places that you find which you can then re-visit later from ‘Your places’ without relying on search to find them every single time. You can also label your home and work address, which helps Maps and Google Now give you estimated travel times when planning a route.

Note that you can’t add labels to real places that are already registered in Google Maps (such as Malls, theaters, offices, etc). For example – When you visit a Mall and want to remember where it is, you can ‘Star’ it to appear under Your Places.  You can’t add a label to that place because it already has a name. On the other hand, let’s say you visit a relative’s house (which obviously doesn’t have their address registered in Google Maps), you can add a Label to that place and name it “Relative’s house” and it will also appear under Your Places.

4. Save Maps for Offline Use

If you’re traveling somewhere and you don’t want to rely on your Internet connection, Google Maps lets you download offline maps and save them.

To save an area for offline use:

  • Search for the area you want to save on your phone.
  • Once you have the map on your screen, tap on the bar at the bottom of the screen and drag it up.
  • Tap on Save and select the area you wish to download using the zoom feature.
  • Give the map a name and it will remain on your device for 30 days before its deleted.

To open your saved maps on your device, you need to access the app options of Google Maps and select ‘Offline areas’. A list of your saved offline maps will appear where you’ll also have the option to open or delete them.

Keep in mind that the downloaded maps don’t include points of interest or navigation — you’re just getting raw map data for the area and nothing more.

5. i) Tap the Compass to Change Your View

Different people prefer different ways of using their maps. Some prefer to always have the map looking North; while others want it pointed the way they’re looking. Thankfully you can switch between the two modes by simply tapping on the compass icon in the bottom-right corner.

ii) Use Advanced Swipes to get Better Views

Google Maps can be manipulated in many different ways. It has a feature that allows us to see virtual building sizes. Tap on the screen with two fingers and drag down to view all the buildings in a certain area. When you want to return to a bird’s eye view, just swipe back up on the screen with two fingers

If you want to rotate the map from its standard view North, move two fingers in a circular motion to get a view down a particular street. You can always tap the compass in the top-right corner to return to the default view.

6. See Where You’ve Been

This feature that has been reserved for the desktop till now, has finally made its way to Android. Google Maps can keep tabs on everywhere you go. Now that may sound pretty creepy, but it’s very useful at times.

Swipe in from the left or tap the three-line icon in the upper left and go to “Your timeline”. Here, you can navigate to any day in the past and see exactly where you went. You can also add notes to certain days to remember what you did.

But if you want to eliminate your tracks, you can just head to Timeline settings and delete your history or turn the feature off.

7. Adjust your Depart or Arrive Time for Public Transit

If you live in some major cities, Public transit is usually a great and much less expensive way to get around the city. But it often takes a little extra planning to get a trip ready.

With Google Maps, you are not restricted to searching for the bus, train and subway times for right now, but you can get information for public transit to leave by or arrive at a specific time.

Once you put in your start and end points in the public transit view of Google Maps, tap the “Depart At” button in the upper-left to launch the time selector. You can then enter a specific time that you intend to depart or arrive at the selected destination, or even choose to just take the last available transit.

8. Add your home and work Addresses:

You can add your home and work locations saved into Google Maps.  Just tap on the three lines in the upper-left corner, choose ‘Your places’ and add the addresses of your home and work. This will make it a lot easier whenever you are out and about to get home quickly because you can tell Google to navigate home instead of always having to type in your own address.

9. Navigate Inside Malls

Not only for roads, Google Maps also works for some malls! If you are near a major mall, zoom in to see the layout of that mall. You can even find some specifics stores inside the mall, find the restrooms, and even navigate through the separate floors.

10. Send Directions to Your Phone

You probably use Google Maps on the desktop as well as your Android phone. If so, there’s no need to start over when you switch from the PC to mobile.

If you’re signed into the same Google account on both your computer and your smartphone, you can just choose “Send to Phone”, and you’ll get a notification that’ll take you to the directions in Google Maps.

Google Maps has been morphing into more of a city guide the past couple of years with suggestions for places to visit and eat. Touch the Explore option from the slide-out menu and you will be presented with lots of different choices for what’s nearby.

12. Add a Pit Stop & Check Gas Prices

This extremely useful feature is a recent addition to the Google Maps app. Now, once you have started navigating somewhere, you can tap the search button to search for another location and add it as a pit stop. Or, if you know you’ll need a pit stop before you leave, tap the three-dot icon in the upper right and choose “Add stop.”

Additionally, if you search for gas stations, it’ll show you the gas prices at different locations so that you can save a few dollars on your pit stop. As for now, it only supports adding one pit stop at a time.

13. Check Your Speed & Speed Limits

This feature actually hasn’t made it into Google Maps yet, but you can still add it by downloading the app “Velociraptor – Map Speed Limit” from the Play Store. You need to simply download the app, adjust the settings to your preferences, and then head over to Google Maps.

You will now see a little bubble off to the side that shows your current speed and the speed limit of the road you are on. You can even set it up to alert you when you are driving off exceeding the speed limit so that you don’t get a speeding ticket.

Share this
07 Oct

How to Display Ads Using Matched Content – Google AdSense

As we all know Google has rolled matched content for many publishers now. If you have some decent traffic to your blog, you most probably have matched content approved for your blog/website.

Initially, Google has shown only related articles within your own blog, later they rolled out native ads on matched content. With the increase in demand for Native Ads, Facebook and Google are throwing more light on Native Ads and slowly it will become mainstream.

Everything you Need to know about Matched Content by Google Adsense

How to Display Ads on Matched Content:

By default ads are not enabled on Matched content. There is a small settings that you need to do, in-order to display ads. There have been few questions raised on this topic on forum(thread), so I decided to make an article on this to make things clear.

  • While creating matched content ad unit, within the settings you will find an option to monetize the ad unit with display ads from other sources.
  • You have to enable it. Once you enabled it, Google will start rolling native ads within your widget along with your related content.


  • Currently, we are noticing about 2-3 ads per unit, but I think with time Google will start showing more ads with better CTR.

Performance of Matched Content – Native Ads(Via Google AdSense):

As of now I see this ad unit is not performing much. But, this is helping in increasing the pageviews and acting as an additional revenue. I encourage you to start showing up Matched Content widget on your blog.performance-of-matched-content-ad-unit

If you are wondering how to display matched content on your blog – follow this article. Any questions on the same, do raise a thread on our forum.

Share this
01 Oct

Do You Know What 7 World’s Biggest Companies Serves For Lunch?

There are billions of companies around the world in which almost every job comes with a few food-related perks, from absolute basics like a fridge where you can store your lunch. Also, company cafeterias and canteens are different all over the world where they can compete with the world’s best restaurants. The cafeteria is important as it’s the place where employees go to take a break. So, to ensure that break is worth everything, these companies serve the best of food, in some of the best of ways.

Have a look what the employees of Top Companies get for lunch.

1. Apple:

The dining area at Apple is called Caffe Macs and is very spacious, light and totally free. They serve Mexican, Italian, Japanese, Spanish and French cuisine at lunch. Breakfast consists of strawberry French toast, pancakes, ju

2. Google:

Google’s cafeteria is one of the world’s best cafeteria. It has several sections of various cuisines, fast food, snacks, frozen yogurt and drinks, and there’s literally something for everyone.

3. Facebook:

It has an Epic Cafe which serves American and Asian foods and can also take a takeout meal. It provides three meals a day, five days a week. The food for both the employees and the office guests is for free.

4. Pixar:

Pixar’s cafeteria is called Cafe Luxo and has huge sized statues of Buzz Lightyear and Woody at the entrance.

This looks like a museum and serves a variety of dishes, such as salmon in maple syrup, pasta with tofu, hamburgers, fried ravioli, steaks, burritos, pizza, and all possible kinds of desserts.

5. Dropbox:

Dropbox, which is the file hosting service located in San Francisco, California, has a cafeteria called the Tuck Shop. It occupies about 400 meters and there are chefs who serve you whatever you’d like to have!

6. Twitter:

Twitter is an online news and social networking service has a cafeteria called @birdfeeder. The sections and menu at Twitter’s cafeteria are named after hashtags such as #comfort food, #tenderloin.


7. Storm8:

Storm8, a mobile social game developer provides unlimited snacks and drinks, catered lunches, dinners including sushi and steaks, and special treats like pork belly burgers from Big Chef Tom of Food Network Star.

If you want to enjoy the food,  go get a job in these World’s top companies.

Share this
21 Jul

Google reportedly working on headset that mixes augmented and virtual reality

Even though Google reportedly scrapped plans for an Oculus Rift competitor, Engadgetreports that it’s still working on a high-end standalone headset — one that mixes features of augmented reality and virtual reality.

According to sources who spoke with Engadget, the company is moving ahead with a dedicated headset that won’t require the use of a mobile phone or computer to function.Engadget says while the headset has a screen, its feature may lean more on augmented reality. It’s not clear what that would look like yet, but it could mean the device is closer to Microsoft’s HoloLens than something like the Oculus Rift. According to Recode, the original device had been designed to compete with other standalone gaming VR devices, but noted that the company had other hardware projects in development.

Google already has some ambitious plans for virtual reality. Earlier this year, it unveiled Project Daydream, an Android-powered VR platform that built on the successes of Google Cardboard. But the company’s long term plans will move beyond Android-powered experiences to include more substantial hardware, according to Engadget.

Google has worked with augmented reality in other areas: it has backed Florida-based Magic Leap, a company creating “mixed-reality” technology, which reportedly includes its own headset. Given the success of Pokémon Go in the last week, it’s easy to believe that the public is primed for a future with augmented reality.

Share this
20 Jul

Google Dabbles in Post-Quantum Cryptography

Google last week announced an experiment with post-quantum cryptography in Chrome. A small fraction of connections between Google’s servers and Chrome on the desktop will use a post-quantum key-exchange algorithm in addition to the elliptic-curve key-exchange algorithm already being used.

The idea is that large quantum computers — if and when they’re built — might be able to break current security algorithms retroactively, so it would be wise to develop algorithmic proof against such cracking efforts.

The experiment employs the New Hope algorithm, which Google considered the most promising post-quantum key-exchange among those it investigated last year. Its aim is to gain real-world experience with the larger data structures post-quantum algorithms likely will require.

Layering the post-quantum algorithm on top of the existing algorithm allows the experiment to proceed without affecting user security, Google said.

Google pledged to discontinue the experiment within two years, emphasizing that it did not want to establish its selected post-quantum algorithm as a de facto standard.

Digging Deeper

“Google’s investigating the quantum computing resistance of New Hope for a robust key exchange algorithm,” noted Rod Schultz, VP of product at Rubicon Labs.

Its announcement “doesn’t herald anything new, but it goes further to confirm that quantum computing-resistant algorithms will provide significant competitive advantage for anyone who has the IP for them,” he told TechNewsWorld.

“You can view this investigation as [one] in Google’s core competency, and also as a hedge and insurance policy around the catastrophic impact to encryption that quantum computing is predicted to have,” Schultz suggested.

The experiment might be putting the cart before the horse, however.

“I doubt that we can develop a defense that works before we actually have quantum computers, because there’s no way to actually test something against a platform that doesn’t exist,” observed Rob Enderle, principal analyst at the Enderle Group.

“Still, this approach could be better than existing methods, making it worthwhile to attempt,” he told TechNewsWorld.

The Quantum Computing Arms Race

There will be a “frantic superpower race to build a quantum computer,” predicted Rubicon’s Schultz.

A bulked-up QC “could undermine the very foundation of modern security by breaking what were once considered unbreakable asymmetric keys in just minutes,” he warned.

There will be a rush to harness this power, if it’s even possible, Schultz said, followed by “an attempt to lock down the knowledge to those who the world thinks will be responsible with this knowledge.”

Post-quantum cryptography is of interest to pretty much everyone on both sides of the law.

“Cybercriminals and government-sponsored organizations are looking at this technology too,” observed Jim McGregor, a principal analyst at Tirias Research.

“No one in the industry believes that any software solution is unbreakable,” he told TechNewsWorld.

Interest in Post-Quantum Crypto

Cryptographers for years have been interested in post-quantum crypto. The seventh international conference focusing on the topic took place in Fukuoka, Japan, earlier this year.

The United States National Security Agency early this year published a FAQ on implementing post-quantum crypto.

The U.S. National Institute of Standards and Technology this spring published a report on post-quantum crypto, and announced an open collaboration program with the public to develop and vet post-quantum crypto algorithms.

Building on years of research, Microsoft this spring established the Lattice Cryptography Library.

IBM this spring made quantum computing available to select members of the public with the IBM Quantum Experience.

Feasibility of Deployment

“Gaining access to powerful computing resources is not difficult anymore,” Rubicon’s Schultz remarked. “The bigger challenge will be in updating the current technology that’s prolific today with QC-resistant technology. It will only take a single quantum computer in the hands of the wrong person to destroy the foundation of encryption today.”

Rolling out post-quantum crypto technology “will likely be coordinated with advancements in the systems used within the data centers,” Tirias’ MacGregor suggested. “It shouldn’t be cost-prohibitive, but widespread usage could take many years.”

However, “Once we have working quantum computers,” noted Enderle, “we’ll use them to encrypt as well as decrypt, making this solution obsolete.”

Share this
18 Jun

Google’s developing its own version of the Laws of Robotics

AI artificial intelligence neural networks

Google’s artificial intelligence researchers are starting to have to code around their own code, writing patches that limit a robot’s abilities so that it continues to develop down the path desired by the researchers — not by the robot itself. It’s the beginning of a long-term trend in robotics and AI in general: once we’ve put in all this work to increase the insight of an artificial intelligence, how can we make sure that insight will only be applied in the ways we would like?

That’s why researchers from Google’s DeepMind and the Future of Humanity Institute have published a paper outlining a software “killswitch” they claim can stop those instances of learning that could make an AI less useful — or, in the future, less safe. It’s really less a killswitch than a blind spot, removing from the AI the ability to learn the wrong lessons.

atlas upgrade 2

The Laws are becoming pretty much a requirement at this point.

Specifically, they code the AI to ignore human input and its consequences for success or failure. If going inside is a “failure” and it learns that every time a human picks it up, the human then carries it inside, the robot might decide to start running away from any human who approaches. If going inside is a desired goal, it may learn to give up on pathfinding its way inside, and simply bump into human ankles until it gets what it wants. Writ large, the “law” being developed is basically, “Thou shalt not learn to win the game in ways that are annoying and that I didn’t see coming.”

It’s a very good rule to have.

Elon Musk seems to be using the media’s love of sci-fi panic headlines to promote his name and brand, at this point, but he’s not entirely off base when he says that we need to worry about AI run amok. The issue isn’t necessarily hegemony by the robot overlords, but widespread chaos as AI-based technologies enter an ever-wider swathe of our lives. Without the ability to safely interrupt an AI and not influence its learning, the simple act of stopping a robot from doing something unsafe or unproductive could make it less safe or productive — making human intervention a tortured, overly complex affair with unforeseeable consequences.

google-car-hed-2-640x353Asimov’s Three Laws of Robotics are conceptual in nature — they describe the types of things that cannot be done. But to provide the Three Laws in such a form requires a brain that understands words like “harm” and can accurately identify the situations and actions that will produce it. The laws, those simple when written in English, will be of absolutely ungodly complexity when written out in software. They will reach into every nook and cranny of an AI’s cognition, editing not the thoughts that can be produced from input, but what input will be noticed, and how will it be interpreted. The Three Laws will be attributes of machine intelligence, not limitations put upon it — that is, they will be that, or they won’t work.

This Google initiative might seem a ways off from First Do No (Robot) Harm, but this grounded understanding of the Laws shows how it really is the beginning robot personality types. We’re starting to shape how robots think, not what they think, and to do it with the intention of adjusting their potential behavior, not their observed behavior. That is, in essence, the very basics of a robot morality.

Google's latest self-driving car prototype (December 2014)

Should this car notice if its engineers start ending its.

We don’t know violence is bad because evolution provided us a group of “Violence Is Bad” neurons, but in part because evolution provided us with mirror neurons and a deeply-laid cognitive bias to project ourselves into situations we see or imagine, experiencing some version of the feelings therein. The higher-order belief about morality emerges at least in part from comparatively simple changes in how data is processed. The rules being imagined and proposed at Google are even more rudimentary, but they’re the beginning of the same path. So, if you want to teach a robot not to do harm to humans, you have to start with some basic aspects of its cognition.

Portal RobotsModern machine learning is about letting machines re-code themselves within certain limits, and those limits mostly exist to direct the algorithm in a positive direction. It doesn’t know what “good” means, and so we have to give it a definition, and a means to judge its own actions against that standard. But with so-called “unsupervised machine learning,” it’s possible to let an artificial intelligence change its own learning rules and learn from the effects of those modifications. It’s a branch of learning that could make ever-pausing Tetris bots seem like what they are: quaint but serious reminders of just how alien a computer’s mind really is, and how far things could very easily go off course.

The field of unsupervised learning is in its infancy today, but it carries the potential for true robot versatility and even creativity, as well as exponentially fast change in abilities and traits. It’s the field that could realize some of the truly fantastical predictions of science fiction — from techno-utopias run by super-efficient and unbiased machines, to techno-dystopia run by malevolent and inhuman ones. It could let a robot usefully navigate in a totally unforeseen alien environment, or lead that robot to slowly acquire some V’ger-like improper understanding of its mission.


Share this
17 Jun

Google wants to use wireless to bring gigabit Wi-Fi to more fiber customers

Google Fiber

Google Fiber has won accolades across the country for delivering gigabit speeds at prices that leave companies like Comcast grinding their teeth in fury as profit margins erode out from under them. One problem that Google and its rivals both face, however, is the expense of running fiber to each individual home. During Alphabet’s (Google’s parent company) annual shareholder meeting, CEO Eric Schmidt said that point-to-point connections can now be deployed via wireless at comparable speed to wired infrastructure, while being “cheaper than digging up your garden.”

Google has previously applied to the FCC for permission to test millimeter-wave wireless networking devices, which typically operate in the 60GHz band. The 802.11ad wireless networking standard also supports 60GHz frequencies, and it’s not a stretch to think that Google might want to build a network using that standard, especially since its already available in routers you can buy today.

It’s not clear, however, if 60GHz spectrum can be easily adapted to real-world conditions. The 60GHz spectrum is largely unlicensed and free from interference, but it’s also severely affected by attenuation from a number of sources. Part of the reason this has such an impact on 60GHz signals is because they resonate with the O2 molecule –otherwise known as “The stuff we breathe” and “21% of the atmosphere by volume.”


Graph by Wikipedia

60GHz signals are also subject to attenuation from rain, foliage, walls, and the human body. There are upsides to this situation, since it would allow for spectrum re-use over a relatively short area, but it’s difficult to see anyone building a cost-effective Wi-Fi network using 60GHz technology. The 802.11ad routers you can buy today use 2.4GHz, 5GHz, and 60GHz bands precisely so that the router can switch to 2.4GHz or 5GHz if you walk out of the room. 60GHz Wi-Fi also requires line-of-site transmission, which could mean Google Fiber would need specialized hardware in consumer’s houses in order to ensure a strong signal.

There are ways to deal with some of these problems, such as using a higher-power transmitter, or installing backup equipment in the 5GHz or 2.4GHz bands that would kick in if the 60GHz signal became too weak and began to fail. The problem in these cases is that people would rapidly sour on the idea of Google Fiber if it turned out they weren’t getting the speeds they thought they paid for. Building more transmitters closer to their customers would also help the situation, but at a higher cost.

Research into the 60GHz band is still ongoing, but some of the problems facing it are intrinsic to the frequency and its characteristics. Better technology isn’t going to solve the line-of-sight problem or the rain fade issue.


Share this
25 May

Google unveils Daydream to create a VR ecosystem for Android

Google Daydream launched at Google IO 2016

Google’s VR efforts have come a long way since two years ago when Google introduced its inexpensive phone-base VR viewer, Cardboard. Today, as part of the IO Keynote, Google’s Clay Bavor previewed Google Daydream — the company’s upcoming VR platform. Bavor stressed the need for a systems approach to VR, especially as it relates to reducing latency — often called the Motion to Photon time. Daydream isn’t a specific piece of hardware or software, but a set of reference designs and Android enhancements that are aimed at creating a vibrant VR ecosystem on Android devices.

Look for Daydream-Ready Smartphones, with support in Android N

Gogle will be publishing the specs for smartphones that it believes are sufficient for a good Daydream VR experience. Those include requirements on the sensors, display, and compute power of the SoC. Most of the major phone vendors are already working with Google on Daydream-Ready devices, and Google expects them to start coming to market this fall. One interesting note, though, is that Daydream is designed to achieve a latency of under 20ms. That is much slower than desktop VR companies consider acceptable for either comfortable viewing of interactive content or action gaming. HTC and Oculus both push for 11ms (providing a 90fps frame rate). Obviously, they also require a lot more GPU horsepower, but it will be interesting to see how many experiences will work in the slower 50fps world of Daydream, and how much discomfort may result.

Android N will include system support for low-latency, as well as a VR system UI, which will help avoid the problem with smartphone-based VR today, where you need to keep going back and forth between VR apps and the Android UI on the phone screen.

Headset & Controller

Googles reference VR headset designGoogle isn’t announcing a headset, but is making available a reference design for headsets. The sketch they showed (included to the right) looks a lot like Gear VR. Some Daydream-capable headsets are expected to be in the market by Fall. The controller reference design looks like a typical Bluetooth remote, but in addition to a button and a touch-sensitive pad also has an orientation sensor like a Wiimote. As you’d expect, you can therefore use it a bit like a magic wand to control your VR experience.

VR Apps & Ecosystem

Google Play for VR will allow users to find, install, and launch VR apps. Your VR apps will then be incorporated into a Daydream Home screen, that looks very much like the one Oculus uses. It’ll be interesting to see what happens when Oculus meets Google on Android phones — will we have both an Oculus Home and a Daydream Home?

Google is also making a major push to add VR support to its core media offerings. Google Play Movies will allow you to view your Play video content in a Virtual Movie Theater, and Google StreetView will be fully VR-ready — you can already use Gear VR and Cardboard with 360-degree photos in Maps through the StreetView app. YouTube is being rebuilt with VR support, including discovery & playlists in VR, with support for spatial audio.

For those hoping Google would upset the apple cart with a stunning new piece of hardware that would bridge the performance, price and complexity gap between Gear VR and the dedicated headsets like Rift and Vive (like me), that didn’t happen. But Google is certainly making the right moves to provide a vibrant ecosystem for VR content creators and users on the Android platform.


Share this
01 Apr

Google’s Making Its Own Chips Now

Google’s Making Its Own Chips Now. Time for Intel to Freak Out

GOOGLE HAS BUILT its own computer chip. And this won’t be the last.



The Internet’s most powerful company sent a few shock waves through the tech world yesterday when it revealedthat a new custom-designed chip helps run what is surely the future of its vast online empire: artificial intelligence.

In building its own chip, Google has taken yet another stepalong a path that has already remade the tech industry in enormous ways. Over the past decade, the company has designed all sorts of new hardware for the massive data centers that underpin its myriad online services, including computer servers, networking gear, and more. As it created services of unprecedented scope and size, it needed a more efficient breed of hardware to run these services. Over the years, so many other Internet giants have followed suit,forcing a seismic shift in the worldwide hardware market.


With its new chip, Google’s aim is the same: unprecedented efficiency. To take AI to new heights, it needs a chip that can do more in less time while consuming less power. But the effect of this chip extends well beyond the Google empire. It threatens the future of commercial chip makers like Intel and nVidia—particularly when you consider Google’s vision for the future. According to Urs Hölzle, the man most responsible for the global data center networkthat underpins the Google empire, this new custom chip is just the first of many.

No, Google will not sell its chips to other companies. It won’t directly compete with Intel or nVidia. But with its massive data centers, Google is by far the largest potential customer for both of those companies. At the same time, as more and more businesses adopt the cloud computing services offered by Google, they’ll be buying fewer and fewer servers (and thus chips) of their own, eating even further into the chip market.

Indeed, Google revealed its new chip as a way of promoting the cloud services that let businesses and coders tap into its AI engines and build them into their own applications. As Google tries to sell other companies on the power of its AI, it’s claiming—in rather loud ways—that it boasts the best hardware for running this AI, hardware that no other company has.

Google’s Need for Speed

Google’s new chip is called the Tensor Processing Unit, or TPU. That’s because it helps run TensorFlow, the software engine that drives the Google’s deep neural networks, networks of hardware and software that can learn particular tasks by analyzing vast amounts of data. Other tech giants typically run their deep neural nets with graphics processing units, or GPUs—chips that were originally designed to render images for games and other graphics-heavy applications. These are well-suited to running the types of calculations that drive deep neural networks. But Google says it has built a chip that’s even more efficient.

According to Google, it tailored the TPU specifically to machine learning so that it needs fewer transistors to run each operation. That means it can squeeze more operations into the chip with each passing second.

For now, Google is using both TPUs and GPUs to run its neural nets. Hölzle declined to go into specifics on how exactly Google was using its TPUs, except to say that they handle “part of the computation” needed to drive voice recognition on Android phones. But he said that Google would be releasing a paper describing the benefits of its chip and that Google will continue to design new chips that handle machine learning in other ways. Eventually, it seems, this will push GPUs out of the equation. “They’re already going away a little,” Hölzle says. “The GPU is too general for machine learning. It wasn’t actually built for that.”

That’s not something nVidia wants to hear. As the world’s primary seller of GPUs, nVidia is now pushing to expand its own business into the AI realm. As Hölzle points out, the latest nVidia GPU offers a mode specifically for machine learning. But clearly, Google wants the change to happen faster. Much faster.

The Smartest Chip

In the meantime, other companies, most notably Microsoft,are exploring another breed of chip. The field-programmable gate array, or FPGA, is a chip you can re-program to perform specific tasks. Microsoft has tested FPGAs with machine learning, and Intel, seeing where this market was going, recently acquired a company that sells FPGAs.

Some analysts think that’s the smarter way to go. An FPGA provides far more flexibility, says Patrick Moorhead, the president and principal analyst at Moor Insights and Strategy, a firm that closely follows the chip business. Moorhead wonders if the new Google TPU is “overkill,” pointing out that such a chip takes at least six months to build—a long time in the incredibly competitive marketplace in which the biggest Internet companies compete.

But Google doesn’t want that flexibility. More than anything, it wants speed. Asked why Google built its chip from scratch rather than using an FPGA, Hölzle said: “It’s just much faster.”

Core Business

Hölzle also points out that Google’s chip doesn’t replace CPUs, the central processing units at the heart of every computer server. The search giant still needs these chips to run the tens of thousands of machines in its data centers, and CPUs are Intel’s main business. Still, if Google is willing to build its own chips just for AI, you have to wonder if it would go so far as to design its own CPUs as well.

Hölzle plays down the possibility. “You want to solve problems that are not solved,” he says. In other words, CPUs are a mature technology that pretty much works as it should. But he also said that Google wants healthy competition in the chip market. In other words, it wants to buy from many sellers—not just, say, Intel. After all, more competition means lower prices for Google. As Hölzle explains, expanding its options is why Google is working with the OpenPower Foundation, which seeks to offer chip designs that anyone can use and modify.

That’s a powerful idea, and a potentially powerful threat to the world’s biggest chip makers. According to Shane Rau, an analyst with research firm IDC, Google buys about 5 percent of all server CPUs sold on Earth. Over a recent year-long period, he says, Google bought about 1.2 million chips. And most of those likely came from Intel. (In 2012, Intel exec Diane Bryant told WIRED that Google bought more server chips from Intel than all but five other companies—and those were all companies that sell servers.)

Whatever its plans for the CPU, Google will continue to explore chips specifically suited to machine learning. It will be several years before we really know what works and what doesn’t. After all, neural networks are constantly evolving as well. “We’re learning all the time,” he says. “It’s not clear to me what the final answer is.” And as it learns, you can bet that the world’s chip makers will be watching.

Share this

© 2015-2020 COBRA Softwares Pvt Ltd. All rights reserved.