←back to thread

303 points zdw | 9 comments | | HN request time: 0.002s | source | bottom
Show context
renrutal ◴[] No.44500376[source]
I clicked the story wondering if the speed of light has changed since the late 90s.

Apparently not.

replies(1): >>44500431 #
1. SV_BubbleTime ◴[] No.44500431[source]
It’s still speed of light in a medium, which is not speed of light. Electricity over copper it is 2/3 iirc.
replies(3): >>44500595 #>>44500642 #>>44500891 #
2. deadbabe ◴[] No.44500595[source]
HFT firms have entire infrastructure that runs very close to the speed of light, beating the competition that runs on antiquated copper.
replies(1): >>44501038 #
3. ta1243 ◴[] No.44500642[source]
speed of packets over copper I think is actually faster than fibre
replies(1): >>44501247 #
4. connicpu ◴[] No.44500891[source]
~200,000km/s is the speed of light in fiber optics. Electromagnetic propagation in copper is more like 99% c.
replies(1): >>44501212 #
5. hhmc ◴[] No.44501038[source]
There's no competition that's running on copper -- even competitors without latency sensitivity with still be running over fibre because that's just the baseline infrastructure in datacentres, transatlantic etc.

Of course, yes, the HFT firms will be using also the standard tricks of microwave towers, shortwave radio, weather balloon etc, to beat the fibre route.

replies(1): >>44501331 #
6. layer8 ◴[] No.44501212[source]
The insulating material significantly decreases the propagation speed: https://en.wikipedia.org/wiki/Velocity_factor#Typical_veloci...
7. layer8 ◴[] No.44501247[source]
It seems to depend on the cable type: https://en.wikipedia.org/wiki/Velocity_factor#Typical_veloci...
8. deadbabe ◴[] No.44501331{3}[source]
There’s always competition running on copper, shitty little traders that think they can beat the big firms.
replies(1): >>44501531 #
9. hhmc ◴[] No.44501531{4}[source]
I don't think the switches connecting to any real exchanges support this