Linux Perceptions

El día de hoy luisdans puso material sumamente interesante. Uno tiene que ver con las Redes Sociales (Networking) y otro tiene que ver con un análisis muy a detalle de Windows vs Linux. La verdad es que me pareció sumamente interesante esa presentación, y retomo una de las ideas:

(Todo lo pueden encontrar en su post de hoy: http://blogs.msdn.com/luisdans/archive/2004/04/28/122607.aspx)

The result of the efforts Microsoft puts into solving customer pain, the approach to development, integration and test and the focus on a broad ecosystem is lower TCO and higher business value for you. When asked why customers are interested in Linux, the most popular response is ‘low cost’. In fact, many customers report that their intuition says that Linux will be 2-3 times cheaper than Windows. Why? It seems intuitive that an operating system that is free would also cost less over the long term to maintain and operate.

Microsoft has long felt, however, that there are many factors (such as ‘people costs’) that would be lower for Windows, and that over time, Windows would actually cost less than Linux. In order to test this belief, and show that our investments in engineering would deliver a lower long-term cost profile, Microsoft commissioned an in-depth study with International Data Corporation to study the 5 year total cost of ownership of Windows versus Linux in 5 common workloads.

What the study found was that initial software acquisition costs are a very small percentage of the 5 year TCO, and that the #1 factor is ‘People costs’. And, the study found that Windows has a lower TCO than Linux in all but the Web serving workload. In fact, the savings in some of the workloads – especially security servers and print serving – were surprisingly high. [Note: The web workload does not include application-centric web deployments; people costs were still lower for Windows in the web workload; the low cost profile overall for the web workload amplified the software acquisition cost differences and the fact that most customers run Linux web servers on less expensive hardware than Windows web servers (a situation that we expect to change with IIS 6.0, which will operate more efficiently than IIS 5.0).

It is also important to note that we’re not stating that TCO is the only factor that companies should use when making operating system platform decisions. Instead, we are asserting that this data shows that the investments we make in rigorous engineering and investing in ecosystem development have yielded returns for customers. In other words, lower TCO doesn’t happen by accident. And we encourage customers to question their assumptions about whether low up-front costs in additional workloads (that we didn’t study) will translate automatically to lower costs over time. We strongly expect that Window’s TCO advantage grows along with the complexity of the workload.

Look at the http://Linux site, where the IDC study is posted, to find additional 3rd party TCO data that corroborates the data found in the IDC study.

Notes:
In the case of the web workload, it is important to note that this category covered only basic web serving (static content and basic dynamic content) and that Windows had a lower ‘people cost’ but hardware costs were higher on Windows than on Linux due to the way that most customers deployed their Linux servers in clusters of very low-end, inexpensive machines. Windows-based web servers tended to not be deployed in clusters and also tended to run more workloads per server, which generally translated to larger, more expensive machines. (Note that because of the advances in IIS Version 6.0, that ships in Windows.NET Server 2003, Microsoft expects that we will come out ahead on the simple web workload in the future.)

Microsoft also had leadership positions on a 3 year basis. In fact, the gap on simple web serving was more favorable to Windows.


Comentarios

Deja una respuesta