Sridhar Vembu is founder and CEO of cloud app service provider Zoho.

With the launch of Apple’s iCloud a few weeks ago, it’s clear there are now two competing visions of cloud computing: device-centric and browser-centric.

iCloud embodies the device-centric world view. The cloud is where the the data is stored, and it is the medium through which data is synchronized between various devices. The application, and more specifically the user interface, is written using the native tool suite, and it resides almost entirely in the device. The cloud version of the data is the truth, and devices contain copies of that data.

Microsoft, by the way, has had the same vision of cloud computing for quite a while, although the company seems to have forgotten this was supposed to be its original vision.

Google, Salesforce and my own company, Zoho, on the other hand, have a browser-centric vision of the cloud. The application logic resides in the cloud and the browser renders the client side of the application, augmented on occasion with native applications as needed. The fundamental operating assumption is that as browser capabilities advance, the need for such native apps should go away.

Google’s Chrome OS has an installation option where you can install a web app, but it is really a bookmark augmented with application meta data.

Today, both views of cloud computing have compelling logic on their side. Apple’s approach is built on the assumption that installed apps are necessary and valuable. The payoff you see here is a rich user interface, instant responsiveness of an app using local data — which is important particularly over temperamental wireless connections — and access to built-in features such as the camera, the GPS and the accelerometer.

In the browser-centric cloud, the compelling features are ubiquitous access to the application and data from any browser anywhere, contextual integration across cloud-resident applications and the relatively easy portability of apps across devices.

So which vision is going to prevail in the long term? Could both visions co-exist indefinitely?

In Apple’s scheme of things, the cloud provides data storage and data synchronization as built-in basic services, and third-party application developers would be able to use these APIs on the client side. So any iOS or OS X application could use these APIs to push its own data to the cloud and automatically synchronize it with other devices, provided of course each of those devices also has that same installed app.

For most personal applications, this model is an excellent fit. The key here is the word personal: In the iCloud vision, the primary target is a person accessing an application and its information from a variety of devices they own, each of which has the specific app installed.

A key limitation here is business-oriented applications, where multiple users view and operate on the same data. A second major limitation is contextual integration across various applications, such as the ability for an email app to pull contact information from a customer relationship management system (CRM) to provide context on a particular email message. In the device-centric cloud, cross-application integration has to happen in the client, which I believe will limit it to some very common use cases.

In the browser-centric vision, the primary target is the group-oriented application, where both the application and its information are accessed by multiple users. In this framework, the application is inherently collaborative and the data doesn’t really belong to a single user but to a broader group or an organization. Most business-focused applications, such as customer relationship management systems and project collaboration apps, fall into this category.

Not needing an installed client is a major asset in this scenario, because ensuring that an entire group of users has the right set of devices and all the right apps installed is painful for a company’s IT. Also, since the application logic resides in the server, it is possible to envision really interesting contextual integration scenarios, such as the email/CRM example I cited above.

Of course, it is possible to envision hybrid scenarios, where a collaborative server-resident application still sports a native installed UI for specific added functionality. A typical example would be a CRM installed app synching with the contacts in your phone or a camera-based app that hooks up with optical character recognition to push business cards into CRM. These apps would not supplant the web UI, but augment it.

Granted, working for Zoho, I certainly have a vested interest in hoping for a bright future for browser-centric cloud computing. But my interests aside, I think it’s clear that while Apple’s device-centric approach will find a comfortable market among individual users, companies that move critical business processes to the cloud will eventually flock to browser-centric offerings.

We’ll be exploring the most disruptive mobile trends at our fourth annual MobileBeat 2011 conference, on July 12-13 at the Palace Hotel in San Francisco. It will focus on the rise of 4G and how it delivers the promise of true mobile computing. We’re also accepting entries for our mobile startup competition at the show. MobileBeat is co-located with our GamesBeat 2011 conference this year. To register, click on this link. Sponsors can message us at