SFP+ to 10Gbase-T Adapters - Running at 2.5G?

[VirtualBacon]

I watched this STH Video, and it mentioned that 6 of the adapters are able to run at 2.5, and 5G.

This is not what I expected at all. As I figured that they were either 1G, or 10G.

The question is, does the 10G Switch need to support 2.5G to negotiate this speed? Or will any SFP+ 10G switch work at that speed?

 

[j_h_o]

For the connection to run at 2.5Gbps, the a) switch port needs to support it, and b) the module needs to support it.

If your switch (like mine, heh) only runs at 1/10Gbps, a 2.5Gbps capable module will not allow the link to run at 2.5Gbps.

 

[VirtualBacon]

This post says otherwise - https://www.reddit.com/r/HomeNetworking/comments/etgg0i/_/ffgcofm
I've found a few other posts of people linking up 2.5GbE handoffs to SFP+ 10G ports with no issues

This sounds like something I need to test out

 

[lowfat]

From my comment on the YT video.

To anyone looking @ using these for 2.5G/5G. I'm not sure I'd buy any except the Mikrotik one. With the Ipolex and likely the other Marvell-based ones, you'll get a 2.5/5G link but the module only broadcasts 10G and real world RX performance is pretty bad.

Also I don't know if the Mitrotik actually performs better. I haven't seen anyone test it. The Aquantia AQS-107 supports 2.5/5G but it is EOL so buying a genuine one will be pricey. There are some 2.5G capable regular SPF transceivers. Haven't tested them either.

You can find other discussions here about poor performance w/ these modules.

 

[VirtualBacon]

I've only ever used one once, it was the fs.com module for an old Synology DS1817 that had 10G BaseT ports on the back, and they got so insanely hot in my Dell X1052P it was unreal

The question becomes, at what point is it just easier to throw a 10G NIC in a desktop, and run at the full 10G speed

Past AP's which need PoE (And to my knowledge, you can't do PoE on 10G BaseT) I don't really get 2.5 or 5G

 

[noduck]

I have used the Mikrotik S+RJ10 at 2.5G and 5G link speed. The speed test results (iperf) were unreliable, simple setups would result in the expected bandwidths, but passing through multiple switches would result in lower speed and a lot of resends.

I have since replaced the 5G NIC (which was connected to a Mikrotik S+RJ10), with a full 10G NIC and a DAC. The PCIe slot is only 3.0 x1, so of course the results are not great, but still much faster (6Gb/s vs 4.7Gb/s) and reliable than at 5G (but still a lot of resends).

Aquantia Corp. AQC108 - Mikrotik S+RJ10 - CRS305-1G-4S+:

Receive: 4.71 Gbits/sec 5 resends​

Send: 2.16 Gbits/sec 19324​

Solarflare SFC9120 - 10G SFP+ DAC - CRS305-1G-4S+:

Receive: 6.09 Gbits/sec 44 resends​

Send: 6.00 Gbits/sec 55305 resends​