Recently, Google's Chrome 2.0 was released boasting a 30% faster performance than version 1. According to one ZD Net benchmark test, Chrome 2 may be twice as fast as it's predecessor.
While Chrome does have some other feature upgrades, they are not earth-shaking, and some would say downright underwhelming. It appears that Chrome 2.0 is hanging its hat on superior performance, and possibly the combination of Chrome plus GoogleApps producing the killer cloud application. Color me cautiously optimistic.
Let's face it, when measuring performance, the one bottleneck that we may ignore take for granted in the Cloud Computing Era is the browser. I'll expound on this in the future, but performance can be classified in three segments: the performance of the network, the performance of the application being used, and the Cloud infrastructure performance. The browser can be considered part of the application performance, but that is dependent on the (sometimes fair) assumption that all browsers - Explorer, Firefox, and Chrome - are created equal in performance. Clearly, we have a challenge to that assumption.
Clean and efficient code in the cloud application is vital of course, but how the browser processes the resource-intensive (and potentially performance-killing) JavaScript is crucial. A suitable analogy would be me is driving my beloved muscle car on a dirt road. Trust me, having a 300 horsepower engine is a beautiful thing. Now imagine trying to drive said car on a dirt road, trying to hit top speed while dodging potholes, rocks, and a possible ditch. On the other hand, a well-paved road does wonders for top speed and a smoother ride. So let's see if it's true - can Chrome provide a high-performance road to the cloud?
Here are the results of some benchmark testing of Chrome. I'm bothered that the metrics aren't readily available (Mental Note to myself: I think that's a future column). But based on relative ratings, Chrome 2 certainly seems to pass muster.
Check this out this graph that charts the relative performance of Chrome 2.0 vs other browsers:
Now I attempted a rather straightforward approach to testing Chrome. In the real world, the end user had a simple bottom line: Does this page load quickly? How quickly I can access the dropdown window?
With that in mind, my test is simple. Test the same three Web 2.0 sites using Firefox, Chrome, and Chrome 2.0
The web applications that will be used? ZipCar will be one choice based on the in-browser application. The second choice will be Facebook, and the various widgets that I can install and upload. The final application? GoogleApps. Each test will feature simple timing tests for 1) Opening the application, 2) Time to load secondary embedded applications, and 3) Upload time for various files (such as pictures for Facebook, or a PowerPoint presentation to Google Apps.
The following results were for IE7, Firefox 3.0, Chrome 1, and Chrome 2.0
Results
Zipcar.com
|
Test |
IE 7 |
Firefox |
Chrome |
Chrome 2 |
|
Open
Site |
13.8 |
11.8 |
10.4 |
7.2 |
|
Load
Secondary Application |
7.9 |
6.4 |
19.9 |
5.5 |
|
Upload
file |
N/A |
N/A |
N/A |
N/A |
Facebook
|
Test |
IE 7 |
Firefox |
Chrome |
Chrome 2 |
|
Open
Site |
19.4 |
14.7 |
21.7 |
8 |
|
Load
Secondary Application |
29.4 |
13.9 |
22 |
16.2 |
|
Upload
file |
20.4 |
20.6 |
23.4 |
13.2 |
Google Apps
|
Test |
IE 7 |
Firefox |
Chrome |
Chrome 2 |
|
Open
Site |
7.9 |
5.3 |
5.7 |
2.9 |
|
Load
Secondary Application |
23.2 |
19.5 |
23.3 |
9.5 |
|
Upload
file |
35.2 |
30.9 |
27 |
29 |
Conclusions:
There is no doubt that Chrome 2 is a decided improvement in performance over its predecessor. In fact, it seems that Chrome 2 can compete with Firefox. But there is a caveat involved here. I ran Firefox with a few extra bells and whistles (as a good consumer would) such as colored tabs and a Facebook toolbar. Chrome 2 was almost stripped to the bare essentials. The relative lack of new features in Chrome 2 is still a point of contention among some. The positive is that less "clutter" should enhance performance. But when I weigh the added functionality of Firefox against the raw speed of C2, I'm willing grant Firefox at least a draw if not better. Nevertheless, I think we're seeing the bottleneck of browser performance being resolved, which means a cleaner path for cloud applications.
Comments