Connect with us


Google Sales Chief Says Desktops Will be Irrelevant in 3 Years



The headline for this post leaves out a salient point in this discussion, and it also hits at another that affects all of us who love our mobile gadgets. The point left out is that what John Herlihy of Google is referring to is that most consumers will be using mobile gadgets to consume information and entertainment.

The translation for that brings up the second point. Herlihy is a Google VP for Global Ad Operations. Obviously is focus is on placing ads on all that content he sees you and I consuming. Of course he’s hoping that will be the case. That’s his job.

His statements go further and he makes some interesting claims. He says that in Japan today most research is done on mobile devices not desktop PCs. As big of a mobile geek as I am, I don’t see this trend accelerating quite that fast, at least here in the US. Technology is advancing at a rapid rate, so rapid that it is almost impossible to keep up. That’s one of the strengths in the industry’s growth and also one of its weaknesses.

The strength is simple to see, even in tough economic times we seem to have an insatiable hunger for these devices we can carry in our hands. But for desktops to really become irrelevant, consumers (and to an extent businesses as well) will need to see something really take hold with some sense of lasting significance before we’ll see desktops start disappearing from dens and cubicles. Call me silly, but I think the breathtaking pace of change at the moment is too quick for things to really permeate with any lasting affect.

Now, maybe 7 years is another story. What do you think?

Via SiliconRepublic



  1. Mike

    03/04/2010 at 11:32 am

    I think he is just wrong.
    Mobility has necessary costs. Chiefly these are power, size of input/output peripherals, and modularity/ease of upgrade.
    Yes a lot of people lead oh-so busy lives and get a lot done with mobile devices in the nooks and crannies of those lives.
    But desktops will not become irrelevant until mobile devices plateau on usable power and tech advancements make human I/O with those devices as good as a full keyboard/mouse.
    It will happen yes…
    But not in 3 years.
    Not in 5.
    Probably not even in 10.

  2. Mike aka C-141xlr

    03/04/2010 at 2:35 pm

    I don’t see this happening at all. They may morph into something much better such as a mobile 3D device that appears where ever and when ever you need it and accepts complete voice input.

    But for now with sci-fi the only place this is happening, we still need our full size keyboards, larger screens, and power. I am not play WoW on a laptop. Maybe if it docked into a system with the other devices, but right now it does not have the power or good enough docking systems.

  3. SAM

    03/04/2010 at 3:13 pm

    Until the horse power and upgradability happen to a laptop,
    the desktops will probably be around for a while longer.

    I can remember in the 70’s the video phone, wrist radio, and
    the 3 day work week would be the norm by the year 2,000–
    so said the experts.

  4. Chris

    03/04/2010 at 3:30 pm

    When anyone makes the claim that the desktop will become obsolete the immediate response is to make a counter claim such as the following: mobile devices lack computing power and input functionality, and therefore, they can not fulfill all the needs of users. I agree with the later part of the claim – that human I/O needs to improve, but I absolutely disagree with the first part…here’s why.

    As it stands right now there is clearly no way that an iPhone (for a popular example) could replace a desktop computer. Why? Because its computing power is limited. There two ways around this.

    1. The traditional approach has been to cram more cpu power into devices, but as we know this leads to shorter battery life. And this is not good because, as we also know, battery technology develops much slower than cpu technology due to problems fundamental to chemistry. Thus, cramming more cpu power into an iPhone leads to a poor user experience – the device dies a couple hours into the day – and is not an adequate solution. But all hope is not lost…

    2. What if, instead of requiring all the computing of a device to occur locally, we perform the computing at some server – a cloud. Then the device must only be responsible for displaying the UI. Allow me to explain further. Consider the scenario in which an operating system exists as a virtual machine on a server (maybe you set this up yourself at your house as a “desktop” or maybe the service is provided by a company – lets say Google). The all you have to do is access the internet, log into your gmail account, and bam you have a full functioning operating system with all your files, all your settings, and which functions like a locally hosted os. The difference, of course, is that your os exists on and is being handled by a server at some remote location.

    Impossible? No…I don’t think so…and maybe Google’s sales chief is giving us some real clues to the near future. What would a system like I describe require. Well first off it would require some script protocol for transmitting a pc like experience over the internet. But this exists (!) and is being increasingly used by large corporations and the health industry. They use “thin clients” which do no local computing, but access a desktop virutalization server for their os’s and data. The technology is unimaginatively named “personal computer over internet protocol” or pcoip. If all of our devices are basically “thin clients” then we don’t really need a full featured operating system. We may just need a very simple os that can connect to the internet. The os could even be browser based – like Chrome OS! The final piece of the puzzle is to figure out a way to get the data to and from the mobile device efficiently and quickly enough that the user experience is exactly (not nearly) the same as a locally hosted os. This would require a superfast, high bandwidth network. But Google is working on this too – just look at Google, Kansas.
    The basic idea of what I describe can be seen in the following video clip from the company Citrix on youtube: Be patient, the good stuff comes around 3:00 minutes.

    There are a couple of obvious downfalls to this system. First, you don’t have a local copy of your documents since they are kept on some cloud. I think there are easy ways around this. There is also a privacy issue associated with this, but I’ll skip it for now. The second, and larger issue, is that we don’t always want to have to be connected to the internet. To solve this a system could be developed to dynamically pass the os state between a local processor and a could processor. This could be done in a way that is similar to Lenovo’s U1 tablet/clamshell concept. A local processor in the tablet is used until the tablet is docked into the clamshell at which point the more powerful chip takes over.

    In conclusion, we often think of our computing experience as having developed linearly. We began with operating terminals which connected to mainframes, progressed to the pc we know and love today, and will eventually have some super devices which compute infinitely complicated material locally. I suspect, however, that our computing experience will be quite cyclical. Thin clients are nothing more than glorified terminals and if this is the way computing (mobile computing) is headed then we have already seen the future – it is in the past!

    Am I crazy for thinking such things? Happy Computing all!

    • Ben

      03/04/2010 at 5:25 pm

      I agree with you that this kind of thing will likely be widespread in the future, but certainly not the near future. Outside of cities (even inside cities, to a lesser degree) wireless internet access is just too slow and unreliable to provide us with an acceptable experience using this cloud+dumb terminal system.

  5. Chris

    03/04/2010 at 3:41 pm

    I should also say that I don’t think the desktop will be obsolete in three years. Some people will still need or prefer a local computing environment, but perhaps desktops won’t be NECESSARY.

  6. Ben

    03/04/2010 at 5:40 pm

    If by “mobile device” he also means laptops/netbooks, then the desktop is already largely obsolete for the majority of people, especially consumers. However, there are and will continue to be for foreseeable future people who need to have a desktop system for work. Phones, on the other hand, are useful but unless something amazing happens in the next 3 years, I doubt they’ll be usable for serious (even non-cpu intensive) work.

    Furthermore, in the home, the desktop won’t really be gone, but I think it’ll change into something else, like a home server. A “desktop” computer still satisfies this scenario nicely.

    • Nameless

      03/05/2010 at 12:35 am

      That’s what I see today’s desktops becoming-tomorrow’s home servers, a sort of mainframe that’s affordable enough for every home.

      And while I do predict more emphasis on cloud computing, sort of like the mainframes of old expanded through the Internet, what I really want to see are more personal clouds. Just one user’s smartphone and various terminals at home to access that one user’s home server, without dependence on anyone else save for whoever provides the Internet connection away from home. No dependence on Google’s or anyone else’s servers.

      The current Remote Desktop paradigm is inefficient, however. The vast majority of the time, it’s not the desktop itself we want-just its applications and its processing power. The server should thus distill its interface to make that easier to access. One proposal of mine would be a “remote application shortcut”-it looks and acts a lot like a shortcut to a local executable, and when you run it, it feels mostly the same, too-only all the heavy lifting is being done on that home server.

      Also, each terminal should have an interface to match, both in form factor and usage patterns. If I were using a smartphone to tap into the home server, chances are I’m using it as cloud storage ala Dropbox. If I’m in the living room, chances are I want multimedia functions. And if I’m in a home simulator cockpit, chances are I want to fire up a fitting vehicle sim in there and do nothing else.

      Whatever the case, desktops won’t disappear, just change in their usage patterns-perhaps so that we don’t have to walk up to them to use them, but use them wherever it makes sense. I certainly know I’m not sacrificing the upgradability, expandability, and raw power only a desktop can offer-modern games aren’t getting any less demanding!

      (And before you bring up console gaming: it’s nice and simple, but it’s far too limiting in comparison.)

  7. Paul Harrigan

    03/04/2010 at 8:23 pm

    I just don’t see every stock trader giving up his Bloomberg for an ipad, every engineer giving up his large-screen high res monitor for a laptop, or every security professional giving up his multi-monitor display system to watch a series of blackberries.

    The use of mobile devices may be ubiquitous for people who already are mobile and for students — i.e., the largest users of the google pages for which this guy sells ads. Search, however, is a small part of the work that people do every day — an important part, but a small part — and this guy is failing to see the world beyond his work.

  8. KaH

    03/04/2010 at 9:16 pm

    The paperless office was prediced 20 years ago – that worked

    Cobol was going to disappear 15 years ago – still being used

    Bill Gates mentioned something about the amount of memory – go it wrong.

    The desktop for many computer gamers and at the work desk will be around for some time to come.

  9. Scott

    03/04/2010 at 9:42 pm

    Desktops will still be out there until we reach a world like that described in “Rainbows End” (

    …even then, some will still demand a desktop supercomputer of some kind.

  10. Medic

    03/05/2010 at 9:37 am

    It is time for the google sales chief to read this comment section.

    I disagree that desktops will disappear in three years time. Although I have a tablet pc and an HTC HD2 pocket pc, I find myself using the desktop most of the time. It is pratical and keeps you focussed on your work, because mobility in a sense requires multitasking, surrounding (environmental) factors distracting you from your work. The problem I experience is that the current user interfaces and connectivity still doesn’t make ease of productivity. It is all still too much fun and play and endless tweaking. Unless all this is a temporary chaos, promising that the current wastful investment will promise that what has been dreamt of in the last 5-10 years, I will certainly shout hooray. In my experience companies are not clear in where they are heading to. It is all micro-economics with a maximum of .5-1year perspectives. What we need is a 3 to 5 year perspective on current built technology. Where is capacative touch going to, what happened to tablet-and-ink technology. What about the promise of further development of current inking technology and improvements of user interface. All seem to have gone down the drain, while it had something good going. And what is with this love for capacative? And what about the combination of resistive and capacative R&D? Companies are not realising it had a good thing going. Sales for capacative and 1 ghz processors may be soaring, but it is not a recipe for predicting the disappearance of dektops in 3 years. Capacative will be a one day fly in 6 months to one year. As is said earlier, the processor power is outgroing the battery improvements. If the speed of a 2 ghz processor cellphone can only be enjoyed at optimum capacity for an hour or two, the fun will die out quickly and people will want to go back to the grid with their desktops. In europe, current restraints on data connections, with fair use policy, cause a poor user experience. This is not good for sales and the evolution of current technology.

  11. Techni

    03/07/2010 at 3:19 pm

    Given the current rate of technological evolution with our wireless cell network, it’ll be 30 years before this is even possible.

Leave a Reply

Your email address will not be published.

As an Amazon Associate I earn from qualifying purchases.