Real world Cat5 performance?

It's easy to find endless articles about the performance of Cat5 vs Cat5e vs Cat6 but I'm curious about the real world performance. I frequently discover that clients of mine will have Cat5 in their homes and offices, but are able to operate at gig speeds anyway. Are folks likely to see a clear advantage in those situations from upgrading to Cat6?

For example, I have a client with a decent-sized network. About 30 workstations, several gigabit switches, two servers, etc. The servers aren't in the network closet with all the rest of the gear (not sure why, before my time), and are instead located about 50' away, connected by a Cat5 cable that's presumably longer than that. I'm tempted to suggest that they at least have me run a Cat6 line from the servers to the network closet. How much would that improve things?

Probably nothing. If they reach gigabit now, going Cat6 won't make much difference. It will if/when they decide to go 10gb though.

Yeah. Cat6 is just a waste of effort/money over Cat5e for gig-ethernet. Either will perform exactly the same, at a full gigabit speed up to 100m distance between hubs/repeaters.

For older Cat5 (not Cat5e), the maximum distance for gigabit will be less (I forget, but something like 30m instead of 100m).

Note that I have actually run 100mbit/sec ethernet over Cat3 telephone wiring, but only for about 50 feet of distance. The shorter the distance, the more forgiving ethernet becomes of cable quality.

If you're trying to use 2.5gigabit or faster, then Cat6 becomes useful.

EDIT: Note that this may depend somewhat on the cable quality despite the cat5 or cat5e designation on the sheathing. The cat5e stuff I use here claims 350Mhz bandwidth, which is better than a lot of cat6 stuff (and not the same thing as the bits/sec speed ratings),

Thanks folks. That's about what I was thinking. I had read that crosstalk is better on Cat6 but I don't know how much effect that would have. I suspect there are plenty of other bottlenecks on their old equipment that would need to be addressed before the wiring becomes the factor.

I will say that it doesn't seem like the nicest wire, TBH. And I do wonder what the MHz rating is. That's not something I've paid much attention to in the past.

A lot of CATxx cabling is not certified or tested to any particular spec by any entity other than the manufacturer themselves. They decide what to print on the cable sheath.

One can hope that the manufacturer of the cable you purchase or find already installed was scrupulous about actually meeting (or perhaps exceeding) the spirit and requirements of whatever claim/specification is printed on the outside.

Cable testing in-situ, with appropriate equipment, can provide some validation of actual cable performance.

I have a Pockethernet test unit, which can be useful at times. I have not used it a lot, but I do have it in my toolkit.

Agreed on all counts. When I buy my bulk cable I try to look for user reviews from folks who have Fluke testers and have done some testing on the boxes they've received.

I actually have a Pockethernet too! I think I might have posted about it on the boards here before, but probably didn't follow up on it. I like it, though sometimes the results are a little puzzling, and the distance to fault readings are almost never useful, TBH.