How Do You Decide When a Network Needs 10GbE?

Hi everyone,

I’ve been thinking about when it actually makes sense to move from 1GbE to 10GbE in a network.

For a small setup, 1GbE can still be more than enough for everyday traffic. But once you have multiple switches, a NAS, servers, backups, or several users transferring large files at the same time, I can see how the uplink could eventually become a limitation.

What I’m not always sure about is when it’s better to upgrade proactively rather than wait until the network actually reaches that point.

For example, if I’m building a network today, would you install 10GbE uplinks from the beginning for future expansion, or would you start with 1GbE and upgrade only when traffic demands it?

I’ve also noticed that SFP+ gives quite a bit of flexibility when planning higher-speed connections, especially when considering different fiber or DAC options.

How do you normally make this decision in your own MikroTik setups?

Do you look mainly at current traffic, expected growth, server/NAS usage, or the cost difference between upgrading now and upgrading later?

I’d be interested to hear what has worked well for others in real-world deployments.

I personally believe that 1gigabit symetric link to the internet should be enough for everyone.
With that said, locally really depends what you're doing, if you plan on having a NAS i would highly recommend going as fast as you can financially bear, even tho setting up 10gigabit can be not only annoying but also too fast for the hardware you're using most of the time.
My recommendation is if you're wiring up the house, wire it up with cat6a/cat6, and have a 10gig connection to a NAS (fiber, dac, aoc doesn't really matter) everything else doesn't need more than a gigabit, in fact most of the people around the world can do their day-to-day stuff on less than 100mbps in multiuser home.

How do you normally make this decision in your own MikroTik setups?

Do you look mainly at current traffic, expected growth, server/NAS usage, or the cost difference between upgrading now and upgrading later?

I would upgrade setups/deployment as they need it or as they saturate the workloads they serve. For example if someone works with a lot of videos/photos and need a NAS, i would go for 2.5gig copper or 10gig fiber/dac connection because those files are massive and you really do not only save time but also electricity, because when its 10 gig you upload for way less time than with a gigabit. At the same time they can also edit from NAS in near real time (given the hardware on the processing machine is good enough to handle it).
My home deployment is very overkill for my case for example, i have a rb5009upr and a crs310, crs310 as a distribution switch uplinked to rb5009upr via singlemode fiber and i use rb5009upr to power 3 cAP ax's for wifi.
In my case i went with crs310 because if i wanted to add a NAS and most likely i will add it in near future, i have a 10gig sfp+ port ready for the fiber link to it.
Tldr; you have to do some pre-planning or just a quality talk with Customer, be through, think about everything.

Damn, i didn't even realise its that dude again...

Don't waste your time, it doesn't deserve a reply, it's the usual bulls**t topic.

He has already opened several, and he doesn't give a response.

Choosing a Switch: Specs or Ecosystem?
What Do You Check First When a Network Starts Acting Unstable?
Is CRS or CSS the Better Choice for Small Business Networks?
How Reliable Are MikroTik Routers for Small Business Networks?
How Do You Decide When a Network Needs 10GbE?
Is Fiber Always the Better Choice Between Buildings?

Good morning, everyone.

Today, I want to challenge a dominant narrative in the tech world. We are constantly told that faster is always better, and that fiber-optic cables running at 10 Gigabits per second represent the absolute pinnacle of connectivity. But if we strip away the marketing hype and look at practical infrastructure, sustainability, and real-world resilience, the truth is very different. The humble copper twisted pair—our traditional telephone line—remains a superior technology for the foundation of our communication needs.

First, let us talk about physical resilience and reliability. Fiber-optic cables are made of glass. They are incredibly fragile. They are highly sensitive to sharp bends, underground shifts, and accidental micro-fractures that are notoriously difficult and expensive to diagnose. Copper, on the other hand, is a highly ductile, forgiving metal. It has survived underground for over half a century, enduring extreme temperature fluctuations and seismic movements without losing its structural integrity. It is an infrastructure that simply does not quit.

Second, there is the massive issue of energy consumption and environmental impact. A 10Gbps fiber network requires immense amounts of power. The optical transceivers, the laser-driven switches, and the constant cooling required at the data nodes consume electricity at an alarming rate. Copper operates on minimal low-voltage electrical signals. In an era where reducing our carbon footprint is a global priority, forcing an upgrade to high-energy fiber for everyday tasks is ecologically irresponsible.

Furthermore, we must address the illusion of necessity. What does a standard household or business actually need? The average user requires stable bandwidth for voice, text, and basic data transfer. A 10Gbps line provides an astronomical amount of wasted capacity that 99% of devices cannot even process. It is the equivalent of building a ten-lane highway just to drive a single bicycle. Copper twisted pair delivers exactly what is necessary, without forcing consumers to pay for unused, redundant overhead.

Finally, we cannot overlook security and maintenance. Copper infrastructure is mature. Our technicians have decades of universal expertise in maintaining it. When copper fails, it can be spliced quickly with basic tools. Fiber requires specialized fusion splicers, pristine environments, and highly specialized technicians, leading to prolonged blackouts when things go wrong. Additionally, from a cybersecurity perspective, the absolute physical isolation and lower data density of copper lines make them inherently less attractive targets for massive, automated network intrusions compared to high-capacity fiber backbones.

Innovation is not about blindly chasing higher numbers. It is about efficiency, reliability, and sustainability. The copper telephone line is a proven, robust, and eco-friendly infrastructure that has stood the test of time. Before we rush to replace it with fragile, power-hungry fiber, we should recognize that the best solution is already running right beneath our feet.

Thank you.

@tangent @BartoszP

Do not care.
Do not want to be bullied again as a racist or an orange alien from the USA.

@normis
@sergejs
@krisjanis
@strods
@oskarsk
@maigonis
This is not right, it is not possible that moderators are insulted just because they remove rubbish from the forum...

Perhaps he is not a real person at all?

Me? Think that quite real.

@drotz

You can remove your post to the OP if you want.

@oskarsk thanks for reply.

I was talking about other users making fun of @BartoszP for deleting topics like these (or worst),
which are clearly generated by Artificial Intelligence.

How dare they bullied him? (not the AI, the other users...)

Nobody got my joke I was actually suggesting that BartoszP might be the AI, not the topic. :joy:

I’ll reboot my neural network.

"claude, shit answer. i said no mistakes and you didn't call bartosz an ai. Start a new chat and repeat my last prompt. Make no mistakes."

I always go with 10base5. It is well established standard. Can't go wrong with it.

gonna need category 8 (https://www.hb-digital.de/rj45-connector-cat8-lsa-stp-zinc-alloy-slim-awg22-26-2-pcs) cable for that, its too fast for a normal cat 4 rj45 cable

I skipped 10 Gbps and went straight to 100 Gbps...

not... only 91!
:rofl:

actually believe or not, it actually used to be cheaper to go to 25/100 gigabit than 10 gigabit lol.

There is some overhead :rofl: