Many media articles constantly talk about how smartphones and tablets will eclipse the desktop PC and that everything will move to the cloud. I believe that these beliefs are erroneous.
Here are three ways that you can structure software:
- Browser-based web application. e.g. Hotmail, which was launched 16 years ago in 1996.
- Client / server application. e.g. Outlook with Exchange Server
- Desktop application. e.g. Outlook without Exchange Server
Each structure has advantages and disadvantages. Email works well in any of the formats above. Some software such as Microsoft Word and Adobe Photoshop work poorly as browser-based web applications. Just try Google Docs or the online version of Photoshop. These products will likely never be comparable to their desktop counterparts since it is difficult to do complicated user interfaces through a web browser. The UI designer has to carefully navigate around the differences/bugs between browsers and browser versions. Another approach would be to use a Java applet (e.g. Yahoo’s stock screener). However, Java applets are slow and not very popular nowadays.
In the future, more types of software will move towards cloud-based delivery and benefit from it (e.g. all users are on the same version, the user doesn’t worry about backups, benefits of software as a service, etc.). However, I don’t see cloud computing dominating all software. Compute-heavy applications such as cutting-edge games will stay on the desktop/console while there will still be less-intensive games delivered through a browser (e.g. Gamestop’s kongregate.com hosts many free Flash games).
The idea that cloud servers will do all the heavy lifting seems far fetched to me. We will not enter a world where people stop buying desktops/laptops because their tablets and smartphones can connect to a server that renders everything including cutting-edge 3-D games. The bandwidth costs would be massive. A server would have to stream full HD with minimal compression to give an experience equivalent to a desktop. This is not technologically feasible or sensible.
Some people seem to have the misconception that virtualization will reduce hardware costs (John Hempton of Bronte Capital seems to have this opinion). The argument is that centralizing hardware onto servers will increase resource utilization. I am not an IT admin but the IT admins seem to be saying that virtualization has similar or even higher hardware costs. One reason is that virtualization slows down performance. The real benefit of virtualization is that it can reduce labour costs (read the linked article which explains it).
I do see virtualization continuing to grow rapidly. Some mid-large business are starting to virtualize all their desktops onto servers, which should cause desktop revenues to decline and server revenues to increase. I believe that virtualization of the desktop partly explains why desktop sales have been slowing while server markets are rapidly growing.
Smartphones and tablets cannibalizing the desktop PC?
I don’t see this happening. Smartphones and tablets have smaller screens. The smaller screen is obviously an advantage (portability) and disadvantage compared to the PC. The touch-based interface is also an advantage (smaller device) and disadvantage (slower typing). They are complimentary to the desktop/laptop PC as a second or third device. I don’t know of anybody who has replaced their desktop PC with a tablet.
I do not believe that the smaller size or touch-based interfaces will replace the desktop or laptop. Smaller size by itself has been tried in the past (e.g. personal digital assistants, netbooks). Touch-based interfaces have also been around for a while.
Social networking, Web 2.0
My opinion is that the new generation of Dotbomb 2.0 stocks are a mini-repeat of the Dot-Com bubble. These stocks are massively hyped and the valuations are ridiculous. A few of the Dotbomb 2.0 stocks like Facebook have viable business models and positive cash flow. These may be reasonable investments at lower valuations. Many of the Dotbomb 2.0 stocks have ridiculously high short interest and borrow costs.
Sometimes correctly spotting the trend correctly will not mean that you make money. We can look back in history and look at the predictions that Bill Gates made in his book The Road Ahead. He was excited about some technologies that never really took off (speech recognition, homes controlled by computers). He correctly predicted that the Internet would change the world. Unfortunately, Microsoft was never able to capitalize on the Internet. It did not have home runs with webTV, Hotmail, or search. I think that similar factors will make many tech stocks equally difficult to predict. While virtualization is a rapidly growing market, I do not know if Citrix (CTXS), VMWare (VMW), Microsoft (MSFT), open source, or some other player will be the winner.
In smartphones and tablets, I think that Intel has a very strong chance of wiping out its ARM competitors (e.g. Qualcomm, Broadcom, Marvell, etc.). I believe that it is one of the few cases where one company has an unusual structural advantage over its competitors.