10Gb/s Ethernet: what I did to get it working in my home

That Mikrotik switch must be heating up like crazy if you fill all 4 ports, since it's using copper .

How about this one? https://mikrotik.com/product/crs305_1g_4s_in

You can use fiber optic SFP+ in it which should not run as hot.

Run fiber everywhere for the fun of it, instead of copper Ethernet. Here is an interesting post about it:

https://sschueller.github.io/posts/wiring-a-home-with-fiber/

You can simplify it and avoid splicing which is the most complex thing about it.

I went a slightly different route. My switches are linked with 10Gbps SFP+ across the apartment, but it was way too late (and too much hassle) to pull proper in-wall fiber. Instead I used one of those ultra-thin (around 0.1mm), unshielded fiber cables, and just snaked it through door frames and taped it along the walls. I'm genuinely impressed this stuff exists, it makes retrofitting fiber into a finished space so much less painful.

Most of my edge devices are still on 2.5GbE though, and I'm increasingly aware that for anything with plain SATA disks, the drives are the real bottleneck. Once I LAG'd 2×2.5GbE to get a 5Gbps pipe, it became obvious the network wasn't the slow part anymore in a lot of cases.

And yeah, the 10GbE SFP+ modules run hot, so hot that I would not lay my fingers on them for more than 2 seconds. I stuck 2 copper heatsinks on my module, not sure they do much but the module runs smoothly. Even so, I'm pretty happy with the overall setup: from my 10GbE-equipped Mac I can saturate multiple machines at once and I no longer think about the network most of the time, which was the goal.

I don't understand the decisions to use RJ45 copper cable for anything higher than 1GbE, unless

  a) You like small computers and these only comes with RJ45 ports and 0 PCIe slots
  b) You enjoy fighting with super rigid cables, port flaps and hot modules

I upgraded my home to full 10Gbps network at 2018. At that time:

  * 10Gbps SFP+ PCIe NICs (Mellanox CX3 and similar cards) are already dirt cheap ($15 for OCP adapted ones, $50 for native PCIe cards) because datacenters are upgrading to CX5
  * 10KM SM SFP+ modules are already dirt cheap (~$6) because 4G BBU-RRU communication uses these
  * Silent 10Gbps SFP+ switches start appearing (still a little pricey); 10Gbps RJ45 switches are not available anywhere

Those 10KM SM modules will happily run off After 2024:

  * 25G/100G SFP+ PCIe NICs are dirt cheap (~$80 for single port 100G NIC)
  * 100G QSFP28 modules are in acceptable price; 25G SM/CWDM SFP+ modules area already dirt cheap (~$50 for 6-slot CWDM ones) because 5G BBU-RRU communication uses these
  * Silent 100Gbps switches start appearing (you can get 24*25G + 4*100G under 100W; still a little pricey); 100Gbps RJ45 switches are not available anywhere

I already have an experimental 6*25G CWDM fiber connection between my 2 rooms. Given that most of my devices wouldn't need anything higher than 10Gbps, I'm thinking of giving each room a single 100G port, split them into 4 lanes in the room with passive wavelength splitters for different devices, which further reduces cost on switches.

TM makes my M2 Air chug when it runs against a local Samsung SSD. It does no such I/O throttling.

Has anyone tried the InvisiLight kits? They give you a 600 micron fiber optical link in the kit which you can hide down the side of carpet etc. I haven't tried it but a few YT vids made it look interesting.

https://www.amazon.com/stores/InvisiLight/page/DFF3C042-0D2E...