Beyond Bandwidth: Why Distance Matters When Choosing a Data Center
Most people think in terms of bandwidth when they talk about network speed. But, since data transfers are constrained by the laws of physics, the distance that the data has to travel is just as important when calculating data transfer speeds.
Round Trip Time (RTT), also known as latency, plays a significant role in determining the speed of the connection. The maximum possible data throughput between two hosts is determined by the TCP window size and the latency as shown by the following formula:
Simply looking at the formula above shows us that the smaller the latency, the smaller the divisor, and thus, the greater maximum throughput. For example, for a typical 64KB TCP window size for a Microsoft Windows machine and a latency of 1 millisecond, you can expect about 524 Mbps of throughput over a loss less fiber connection.
While advances in networking technology have allowed data centers to use more efficient high-traffic routers and fiber optic cables, the reality is that data packets are still constrained by the speed of light (or worse, the speed of electricity for those that use copper cables), making distance a significant factor for ultimately determining data transfer speeds.
In fact to achieve 1 ms of latency (the round trip time) over fiber, the two hosts need to be within 93 miles of each other.
The key takeaway here is that having enough bandwidth, while necessary, is not enough to ensure the performance of remote or cloud-based applications. High latency can negatively affect performance as bandwidth is not fully utilized while applications wait around for responses from distant data centers.
365 Data Centers operates 10 colocation facilities in the U.S. in downtown and mid-town locations. Thousands of businesses within a few hundred miles of our data centers will be able to achieve sub-five millisecond (<0.005 second) latency. Our close proximity to downtown business districts makes it easier to access real-time disaster recovery, backup, and data intensive applications, translating to greater peace of mind, better application performance and a better experience for end-users.