Of course there is one interface less, 4 instead of 5, but for the use I intend this hardware for, I found a couple of possible setups (one using only one router, one using two of them), see starting here: http://forum.mikrotik.com/t/attempting-to-evolve-from-cavemans-failover/170048/1
that both need not more than four ports.
The idea of having the possibility of using the ax lite as an access point later or in another installation, should the performance not be enough in the future is interesting.
Also, the powering via a “normal” USB-C 5V 2.4 A power adapter would probably make it easier to deal with, it seems like these are common for tablets.
Any drawbacks or other possible reasons to still prefer the Hex?
Revisiting this thread, I noticed this bit. Beware that people have had hAP lite ax units blown up by USB-C power adapters putting out voltages higher than 5V. I don’t know if the hAP will use USB-C PD negotiation to get a multi-voltage adapter down to 5V or if it takes whatever the power adapter considers a default voltage or what. A laptop USB-C charger is not likely to put out 5V as its primary design target, for example.
Yep, the amount of possibile confusion with Power adapters is huge but a rather common type has 5 Volts (fixed) and 2.4 or 3 Amperes with 1 or 2 USB-A sockets. These will surely work. In theory a PD charger won’t give more than 5 Volts before negotiating with the device, problem is with non-Pd compliant device that may output 12 or more Volts before negotiating or wihthout negotiating at all, see: https://www.electro-tech-online.com/threads/thoughts-on-usb-c-and-the-pd-standard-noncompliance.165206/
@tangent
Now I see how the misunderstanding came out, I meant “tablet” (or smartphone) power adapter, not “laptop”, the latter ones may have multi-voltage negotiation or anyway can output more than 5 V.
Thank you very much again tangent for your input on the matter, I would have never thought to get an access point instead of a router.
@holvoeth
Yes, I do not need (now) any of the Wi-Fi capabilities right now but if I can have the same (or possibly even better) routing capabilities AND if needed the Wi-Fi, why not?
Thank you very much for the comparison table, it seems that there are no real differences in practice between the two and with the radio on the ax lite disabled it may even need less power, besides the 1 less ethernet interface, the ax-lite seems like more flexible/reusable in the future, and getting them will allow me to experiment/learn about the WI-Fi part.
The fact that it comes already with Ros7 (and with 128 MB storage) should more than compensate in my use case the the misssing USB and SD ports.
As well, I don’t think I have an use for zerotier right now, but it is good to have this option should the need arise.
Since my internet connections are very slow anyway, even if the ax lite it may be slightly slower due to slower CPU and less threads (but maybe this is compensated by the different architecture) the ax lite seems really like a more suitable device.
As a passing comment, the hap ax life is a usefully flexible little device. You are using one for routing, no wireless. I am using two for wireless (AP for Things) with no significant routing (I do keep a separate management network by habit). It is cheap enough that the additional functionality is practically a bonus, good enough that it won’t be obsolete soon.
slower CPU and less threads (but maybe this is compensated by the different architecture)
A useful proxy for determining CPU power in RouterOS devices is to compare the routing test results for the two devices: hEX, hAP ax lite. If you go on straight per-core thread count and MHz, you’d expect around 50-100% improvement in the hEX, but no, it’s more like 20-30%.
My understanding is that there’s a hidden factor at play here as well: the hEX would have been tested under RouterOS 6, not RouterOS 7. MikroTik has yet to go back and re-test everything and republish results for older devices under RouterOS 7. Being simpler and older, RouterOS 6 can be more efficient in some ways than 7, enough that it may wholly swamp that small performance delta.
For me, it’s a more practical matter: 30% speed advantage over nothing is still nothing. If I can’t run my containers, it doesn’t matter how much faster they may in theory run.
