Bandwidth vs. Latency: the business decision

Bandwidth vs. Latency: the business decision

13 August 2020 | Chuck Girt

Cover

Chuck Girt

Blog Author | CTO, Everstream

Cover

How much bandwidth does your company need, and how much latency can it afford? Chuck Girt, Everstream CTO, explores the answers

How much bandwidth does your company need, and how much latency can it afford? The two go hand in hand, and the answer comes down to the applications that are most important to your business. While bandwidth garners the headlines, latency—a measure of how quickly data can travel from one point to another—is perhaps even more important for the real-time applications that are driving today’s enterprises because it has the biggest impact on the customer experience.

Many factors contribute to latency, including packet size, the physical distance between departure and arrival points, the number of hops along the route, and other traffic on the network. By minimizing even one or two of these factors, latency is improved. While there’s no such thing as zero latency, it’s important that when building a network, service providers try to get as close to that nirvana as possible.

One of the biggest applications that require both high bandwidth and low latency is real-time replication of data systems for business continuity and disaster recovery. Some larger enterprises have a thousand bytes of data per second traveling into a database and that needs to be replicated to a backup database in real time. Even a 150 millisecond (two-tenths of a second) delay can lead to poor performance, causing critical data to be missed. In industries such as manufacturing or financial services, that can mean significant lost opportunity and a poor user experience.  

Streaming video also requires low latency. Even before the Covid-19 pandemic brought about the “Zoom era,” large enterprises and international companies have long relied on video conferencing to communicate with constituents such as partners or board members. As more users join the conference, more bandwidth and lower latency is required. High latency can result in the jitter and delays that are all too relatable by many users. Poor user experience can ultimately consume the focus of the call, rather than the original purpose.

Any company can benefit from a high-bandwidth, low-latency network. A few examples include:

  • Remote healthcare. The faster the doctor can get results of tests like MRIs, the faster they can make a diagnosis.
  • Financial services companies. A few missed seconds can mean missed opportunities for any company that is trading.
  • Companies that process large volumes of transactions. Massive databases are used for credit card processing, and even the smallest delay can cause huge losses.
  • Manufacturers. Many manufacturing companies use inventory management systems that track parts and products in real time.
  • Mobile network operators (MNOs). Mobile operators rely on a fiber network for cellular backhaul. Latency can cause broken or dropped calls, which can become a real issue for emergency services involving 911 calls.

How can enterprises achieve both low latency and high bandwidth from their networks? The key is to look at the applications—what are you trying to do, and what is it dependent upon? Some enterprises need a lot of bandwidth and don’t worry about latency. For others, latency is mission critical. By understanding your data needs and working with a service provider that has designed its network to deliver on both latency and bandwidth, you can find the right balance based on your business priorities.  

Chuck Girt is CTO for Everstream