It's a good point. It's not a normal speed test. When my second tech was here...he was waiting on some test results and was on the phone with another tech and an engineer at their Fairfax office. Turns out their "server" was down...as the other tech he had on the phone had the same issue. They swapped servers and were able to get their meter'ed tests back up.
That's why I think their meter Speedtest is similar to the Netgear Nighthawk app (both are actually powered by Ookla Speedtest.net). Which only measures the router speeds to the internet (actual modem)...which are usually clocking 500 Mbps+. It's testing that connectivity to their server but not going through the normal "customer" route.
Try testing on Ookla in a variety of servers nearby, instead of just theirs. Also dslreports as said above. I've had really bad results on theirs recently- yet I'll still get fine speeds from servers states away. cat 6/7 also are irrelevant to speeds. Even with 5e you should (assuming not an incredibly busy environment) be capable of gigabit or near gigabit. Before that, 6+ cables are mostly for shielding.
idk if the tech uses the same ookla test, but cox's servers are just bad and maybe a sideeffect of the rest of the overloading. For reference, though I only have 300mpbs now because 1000/30 is a scam:
it takes around 9 SECONDS to achieve a normalized speed on cox's server about 25 minutes away from me.
It takes two to three seconds to achieve the same curved speed efficiency on a server 2+ hours from me; on multiple different targets.
If it takes so long it can't even set up the test properly, it won't rate as 1000mpbs unless it's able to throughput enough to properly enumerate
something is definitely wrong with cox's local nodes.
Yup. I've tried various servers (besides Cox)..I've tried ones in DC, ones in areas within a 10-20 mile radius, I know which ones are "faster". I signed up for DSLReports, used the tests there and posted there. I've done various speedtests on various sites and its all the same.
The Cox meters are "powered by Ookla Speedtest". So it's Speedtest.net technology, which they've farmed out to companies like Netgear (and they use it on their routers/modems and their Netgear App). But as mentioned above, it's not a true throughput test as it's measuring connectivity to a point..such as "the internet" or that your devices speeds should be 1 Gbps but you're not getting that. For example, your car can drive over 150 MPH to get to the grocery store, but realistically you're only going to max out at maybe 55 mph on the high-way because of traffic. From my understanding, this is why the meter tests are always bias..they say you CAN and SHOULD get that speed...but are never actually factoring the traffic, etc. unless you're on an actual computer (car) going through the "normal" route.
My Netgear app (with the Ookla Speedtest) for my RAX45 router SAYS its connectivity to the internet is 500+ Mbps Down. But it's touching an "entrance point"...not the way a computer or any interactivity device would.
I've taken a break on the internet stuff. Been dealing with it for two weeks since I moved in. But yeah, something is definitely off.
