NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Nighthawk RAX80 Half Speed via Ethernet
Nighthawk RAX80 Firmware: V1.0.1.70. Performance Optimization Data Base v1.278,0.0.21 Nighthawk App Version 2.4.25.962 Modem: NETGEAR CM1000. Internet Provider: Cox Communications with guarenteed...
" . . . . unless you're only talking the speed from your router to your pc."
Yes, that is what I'm talking about. Ethernet cable from CM1000 modem connected to RAX80 router. Ethernet cable from back of RAX80 connected to my desktop. Both cables are CAT 5E.
Speeds tests before cycling QoS Off-On:
- Fast.com = 590Mb
- Speakeasy = 547Mb
- Speedtest = 610Mb
- Cox Gigablast = 628Mb
Just ran, at time of this post 1200EST, following speed tests via my desktop after cycling QoS Off-On:
- Fast.com = 1.0Gb
- Speakeasy = 827Mb
- Speedtest = 878Mb
- Cox Gigablast = 918Mb
I have been able to get FAST.com showing as high as 1.2Gb early in the morning. I do notice a speed drop off from Cox as the day progresses. Cox only claims a 1.0Gb with their gigablast service. Cox represented visited me last fall for a internet issue, and I asked him to test modem and they got 1.1Gb readings for modem directly to desktop, and 1.0Gb with ethernet form modem via RAX80 to desktop.
I find it odd that just by cycling QoS I can get these dramatic differences.
I don't think fast.com is giving you valid numbers. You can't get faster than gigabit speeds from a gigabit connection. And the cm1000 is gigabit without any link aggregation or anything. The only modems that hit faster than gigibit right now are the CM1100 and CM1200 as they have link aggregation.
Not sure why the QOS is causing dramatic changes in speeds.
Did you ever try a factory reset and reinstall?
When/if you try it, don't enable qos, traffic monitoring, access control, or armor/parental controls. See how it works.
Qos isn't recommended at speeds >250mbps.
I've read of settings being retained even through a disable/reboot of a router which is why I said factory reset.