Do You Really Need Multigigabit on Every Port?

Hi everyone,

After reading the different experiences here about 10GbE, NAS setups, and switch uplinks, I started wondering about something slightly different.

Is there really much benefit in having multigigabit connectivity on every port, or is it more practical to concentrate the higher-speed connections where they are actually needed?

For example, a network might have a few servers, a NAS, or an uplink between switches that can genuinely benefit from 2.5GbE or 10GbE, while normal desktops, printers, IoT devices, and other clients may never come close to saturating 1GbE.

I can see the argument for choosing a switch with some higher-speed ports for future expansion, but I’m less convinced that every connected device needs to be upgraded at the same time.

I was also looking at some of the MikroTik switching options with SFP+ connectivity while thinking about possible upgrade paths:

[link removed: mod censorship - advertisment]

How do you normally approach this when building or upgrading a MikroTik network?

Do you prefer:

  • A few high-speed ports for servers and uplinks?

  • Multigigabit on most or all access ports?

  • 1GbE for normal clients and 10GbE only where there is a clear workload?

  • Or choosing a switch with extra capacity simply to avoid replacing it again later?

I’m particularly interested in what people have found practical rather than what looks best on a specification sheet.

For those who have upgraded from 1GbE, what devices actually benefited the most from the additional bandwidth?

And another post opening new topic (which is actually almost identical to one of existing).

@BartoszP @holvoetn Perhaps you should consider to lock this thread and do the same for any new thread OP opens until (s)he starts to communicate bidirectionally in existing topics? Opening post in this thread is bordering to propaganda spam (a link to an online shop posted).

do you want bidirectional communication from a bot? why?

Good morning, distinguished guests and colleagues.

Today, I stand before you to critically examine and challenge a deeply entrenched dogma that currently dominates the technology sector. We are perpetually bombarded with the corporate narrative that acceleration equates to progress—specifically, that deploying fiber-optic networks capable of delivering 10 Gigabits per second represents the absolute zenith of human connectivity. However, if we deliberately look past the aggressive marketing campaigns and instead analyze the raw data regarding practical infrastructure, long-term environmental sustainability, and systemic resilience, a radically different reality comes to light. The traditional, often underestimated copper twisted pair—the backbone of our historic telecommunications network—continues to stand as a demonstrably superior technology for the foundational layer of our global communication architecture.

To begin with, we must evaluate the critical factors of physical durability and operational reliability. Fiber-optic cables are fundamentally composed of glass filaments. Consequently, they are exceptionally delicate and structurally unforgiving. They are highly vulnerable to tight bending radii, subterranean shifts, and catastrophic micro-fractures that remain notoriously difficult, time-consuming, and exorbitant to identify and repair. Conversely, copper is an extraordinarily ductile, resilient, and accommodating metal. It has successfully endured beneath our streets for over half a century, tolerating severe thermal fluctuations and significant seismic activity without compromising its structural integrity. We are talking about an established infrastructure that genuinely refuses to fail under pressure.

Next, we cannot afford to ignore the staggering implications of energy consumption and the resulting ecological footprint. Operating a widespread 10Gbps fiber network demands an immense, continuous influx of power. The sophisticated optical transceivers, the laser-powered switching centers, and the intensive cooling systems required to prevent data nodes from overheating collectively drain our electrical grids at an alarming velocity. In stark contrast, copper infrastructure functions efficiently on low-voltage, minimal electrical impulses. At a time when mitigating climate change and shrinking our carbon footprint are urgent global imperatives, aggressively mandating a transition to energy-hungry fiber networks for routine, everyday digital tasks is nothing short of ecologically irresponsible.

Furthermore, we are obligated to deconstruct the widespread illusion of necessity that surrounds modern bandwidth requirements. Let us realistically ask ourselves: what do the vast majority of households and enterprises actually require on a day-to-day basis? The typical user simply demands a dependable, uninterrupted connection for voice transmissions, text communication, and standard data retrieval. A 10Gbps pipeline delivers an astronomical, excessive volume of data capacity that 99% of consumer hardware cannot even physically process or utilize.

Investing public and private billions into this technology is the exact economic equivalent of constructing a six-lane commercial highway to connect a quiet residential cul-de-sac. It is a monument to structural waste, and it is time we re-evaluate the enduring, sustainable value of the copper network we already possess.

Thank you.

@normis
@sergejs
@krisjanis
@strods
@oskarsk
@maigonis
But where have the forum moderators gone?
Are they all busy deleting posts from those who complain that they're not doing their job?
(whether paid or not, if you don't want to do it, stop being a moderator)

If they don't have the time or inclination, they should back off instead of offending users.

Let me remove these topics...

Closing that one and other topics as requested.

User silenced for some time. Will see what happen next.

Account suspended for now. Posts/Activity are bit to suspicious.

For all interested&uninterested:

Just happend: Rosi tried to reincarnate as Julia with same style of posting. Banned in advance.