Hacker News Reader

Comment by fooker | Hacker News Reader

You'll be happy to know that the current latencies are about two orders of magnitude higher than what going at the speed of light will offer. That makes me think that we will easily see a 10x improvement in latency in 15-20 years.

View on HN

Source to that? My napkin math tells me speed of light across US is ~10ms, my experience tells me latency is ~100ms (certainly not 1s), so that's only one of of magnitude.

You are not going to stream games from the west coast to the east coast. It will be more like a data center in/near your city to your home (..say.. within a 50km radius). Currently, that pathway isn't very well optimized and I routinely see local services having a latency of 30+ ms in the bay area.

I am suggesting this will be significantly more optimized, not how fast you can communicate over long distances.

How do you get two orders of magnitude? Round trip for light is ~150 ms. Using a cable it is slower, so more like 250 ms. Current latencies are only like 2x that.

Even with free space speeds and tunnels through the earth you only get like one order of magnitude.

you think we'll able to communicate faster than light?

No, the poster is saying that latencies are about 100x the theretical minimum latency due to speed of light, and thinks that will reduce to 10x.

I don’t know if these numbers are true, but poster was not claiming faster than light.

There are several reasons why Internet latency is inflated over direct-route speed-of-light latency. Some reasons may change others may not.

1) speed of light in fibre is 2/3c - unlikely to change soon 2) fibre is laid in the ground on a circuitous route - rarely changes but does happen 3) queueing delay - that's often the main culprit, changeable, but hard (see the buffer bloat debate)

Reply on news.ycombinator.com