# Speed benefits when switching LAN from 1Gb copper to Fiber Optic

We are trying calculate how will the network performance improve when we switch from 1Gb copper LAN to Fiber Optic.

We have 5-20 people simultaneously working on a Server via the 1Gb LAN. There are constant read-write operations since we work with graphics, e.g. image sequences, large video files, etc.

The real transfer speed for a single user on a good day is around 50MB/s (0.4 Gb).

We would like to know the more or less average increase in speed (for a single user as mentioned above) should we switch to Fiber Optics. I looked everywhere for any real-life values but there are only theoretical numbers.

Thanks!

• which fiber optic speed do you consider? 1/10/40/100Gbs?
– JFL
Commented Nov 21, 2017 at 15:13
• Copper and fiber running at the same speed run at the same speed. So 1 Gbps copper and 1 Gbps fiber run at the same speed. The main reason to use fiber for any connections is for length. Commented Nov 21, 2017 at 15:18
• The length is not so much of an issue now. Let's assume we go for 40 Gbs. Commented Nov 21, 2017 at 15:34
• What's the configuration of the server, esp. the drives? To push enough data to even saturate a 10 Gbps link, you'll need SSDs, and several of them. As long as you're under the practical max (around 700-800 Mbps)) of what can be moved over 1 Gpbs link, your bottleneck is elsewhere than the network. Commented Nov 27, 2017 at 9:25
• Did any answer help you? If so, you should accept the answer so that the question doesn't keep popping up forever, looking for an answer. Alternatively, you can post and accept your own answer. Commented Jan 5, 2021 at 2:41

You are conflating many things here, so let's try to detangle the issues in your question.

1. Data rate is data rate, regardless of the physical medium. A 1Gb connection has the same data rate whether it is fiber or copper.
2. As @toddwilcox mentions, the advantages of fiber over copper are longer spans and electromagnetic isolation. Data rates are independent of the medium.
3. Transfer rates depend on every link in the chain from source to destination. You can only go as fast as the slowest link. So to determine if a user's transfer rate will increase, you have to consider all of the components from the server to the user's computer.
4. You may get a modest performance boost by fixing one bottleneck, but then you will run into another. For example, you may increase your network bandwidth, but you might then be limited by the user PC NIC, disk transfer rate, CPU, etc. The same restrictions may apply to the server too.

You mention you get 0.4 Gb "on a good day." Have you tested when only one user is active? If the number is still in the same range, it probably isn't a network problem.

You also don't mention the kind of network equipment you have. Some consumer-grade equipment can have much poorer performance than their port speeds would indicate. In other words, they may have 1Gb ports, but be unable to forward traffic at anywhere near that speed (I should also mention that consumer-grade equipment is off topic on this forum).

Working with network shares, the #1 bottleneck is usually the server, not the network. For 20 users each simultaneously moving 400 Mbit/s, the server would have to cope with 8 Gbit/s which is quite a load.

In contrast, using a fairly cheap NAS with four large HDDs would work well for a single user, OK for two users, but it would totally cave in with 10 or even more heavy users: only four spindles can't handle the required I/Os - you'd need a larger count of HDDs or (much better) SSDs.