Ten years ago, the average x86 (Intel or AMD CPU based) server was only realistically capable of sending or receiving data at approximately 100 MB/s (megabytes per second) externally. Over the course of the past decade, many advancements have been made that allow for much higher levels of bandwidth. In fact, benchmarks ran in early 2007 showed that at least one system was capable of 7,964 MB/s, which is nearly 80 times faster than just ten years ago. Looking at the roadmaps of both servers and server technologies, I expect that these results may more than double by the end of 2008.
In our society today, we have become ravenous for increased bandwidth. The entertainment industry is a prime example of this phenomenon as we're currently in the midst of an explosion of movie, music, and video downloads. Many believe that the Internet will become the vehicle for most entertainment downloads/purchases in the future.
As our bandwidth needs increase, I believe that it is worthwhile to explore the foundations of bandwidth, which lies in the server system itself. This blog will explore the external bandwidth limits of today's server systems. By external bandwidth, this is simply referring to the bandwidth that can be moved into or out of a server platform. In some cases, we might look at specific technologies, or more importantly we'll simply investigate the aggregate bandwidth potential of a variety of systems.
I intend to keep this site informational and talk about the dynamics of server bandwidth as much as possible. Also, I'd like to use it as a vehicle to publish various benchmarks as they pertain to high bandwidth server applications.
As with most things in life, this is not always a straight forward endeavor, but I will always attempt to explain my findings and reasonings whenever possible.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment