Gigagbit unmanaged switch speed shootout? Please.

[dallas7]

All the latest and greatest 1000 Mbps routers, having been designed on planet FourPorts, have only, well, four ports.

Tests here report routing performance ranging from 300 (and less) to 1300 Mbps, even in different models from the same manufacturer. I'd like to see the same LAN-LAN tests run on a handful of unmanaged 8 port switches, too. What a waste it would be to spend money on a 300 Mbps switch when a similarly priced unit might triple that for us reallySOHO builders.

Netgear GS608
Linksys SE2800
Belkin F4G0810
Asus GigaX 1108

D-Link and TRENDnet each have four models and TP-Link two. Selection of the one model to test should be determined by which the manufacturers' claim is their fastest upon interrogation by Dr. Higgins. Although more than one model from each would be nice.

D-Link: GO-SW-8GE, GO-SW-8G, DGS-108, or DGS-1008G
TRENDnet: TEG-S82g, TEG-S81g, TEG-S80G, or TEG-S80Dg
TP-Link: TL-SG108 or TL-SG1008D

A one-time shootout of this magnitude would be so helpful. Speed is the need; management, enterprise-class features and "green" is irrelevant. Unless I'm way off base in my understanding and there's "no difference" in performance between all these choices.

What say?

Thanks!

 

[thiggins]

There is no practical difference in unmanaged Gigabit switches. Buy on price, brand preference and warranty.

 

[dallas7]

Wow. That was fast.

Given a 3-4x speed diff in LAN-LAN for the switches in these routers you test, how would that not be in various stand-alone switches?

Thanks again.

 

[thiggins]

I do not test LAN-LAN performance.

 

[dallas7]

My bad. I thought your tests of NAS file transfers would infer a LAN-LAN performance and I mixed that up in my mind with your WAN-LAN/LAN-WAN tables. I should have posted that up as "LAN-LAN."

I was mistaken in thinking there would be practical difference in choice for, say, an environment for two or more NAS boxes connected to a switch.

Thanks for your time.

BTW, I chose a Netgear RO-318 a few cyber decades ago based on your review. I used it up until a few weeks ago when it was determined it didn't work too well with a Cisco DPQ3212. I'll find another home for it.

No rely expected. Cheers.

 

[stevech]

Am I wrong that some low cost routers have a LAN interface that allows the software to force some ports to not switch, but rather feed into the firmware - to allow Port based QoS and the like?

That would of course affect packets per second if the destination is on the same LAN based on matching MACs done during ARP table building.

 

[coxhaus]

I generally go by the larger switch has a bigger backplane. With larger switches you need to make sure they are quiet unless you are putting them out of the way in a closet somewhere.

 

[PrivateJoker]

I generally go by the larger switch has a bigger backplane. With larger switches you need to make sure they are quiet unless you are putting them out of the way in a closet somewhere.

That's a good point @coxhaus, a lot of the newer switches and their more green efforts (like auto sensing link length and adapting power for each link, turning off ports not in use, etc) there are many fanless 1U switches out, I have a DLink unmanaged that is cooler to the touch than my small Netgear 5 port FWIW.

If you plan on pulling cable yourself, looking for logical places to run wires up to attic above/basement or crawl space below, etc certainly is helpful.