Is there a Next Revolution in Internet Speed?

I

InnocentNoobie

Limp Gawd
Joined
Oct 31, 2008
Messages
231
As a teenager I enjoyed playing games such as the original Warcraft and Ultima Online through a dial-up connection. Later on in life, cable and DSL internet became options, and I, for the most part, have been able to primarily enjoy FPS games online because of it.

I do remember playing duke nukem 3d through dial up back in the day, but the lag that you would experience made it kind of silly. You just simply couldn't take it too seriously, what with people warping all over the place and only 75% of your shots actually registering. It was still fun, but the limitations of internet speed made RTS or MMORPG much more attractive to play online.

I have not heard of another revolutionary technology that would be similarly significant for online gaming as the switch from dial up to 'broadband' (cable or internet) was, but I think it's a fairly interesting thought.

I do know that while living in the dorm at my university, I could ping under 10ms to most servers (good servers) on my half of the U.S. It makes a hell of a difference.

What about technology that would let online FPS gamers play with each other from overseas? As of now, it's amazing that someone in China or Australia can get a sub 200 ping on U.S. server, but that's still too high for the vast majority of gamers to make it relatively enjoyable.
 
As for residential internet speed, the US sucks balls. South Korea, Japan, and Europe have us beat by miles.
 
to an extent we could have 1Terabit lines from here to japan and itstill wouldnt really help much, electricity, light over fiber or whatever the medium still has to travel the distance, therefore you're always going to have distance latency. this is simple physics, electromagnetic energy only travels so fast....believe it or not our tiny little planet still will induce some inherent latency just because of the distance involved.
 
Yup, it will come to a point where it gets as good as it can get. Even if you hooked up your computer in China directly to the server in the US over Fibre or whatever medium you want, it still takes time from one point to get to another. It also times time to register that data being sent and received. There are overhead for programs and games. Now add in the fact that you won't be hooked up directly to the server, there will be a ton of other people using the internet, hell even the people around you using the same bandwidth. The difference between dailup and DSL/Cable is night and day, but from DSL/Cable to the next revolution will most likely not have the same impact for gaming. Your pr0n downloads though, that's another story, 10Gb standard anyone? ;)
 
In the US? You will not see the next speed jump for a long ass time, reason. . . . .Competition, ISP's have squandered billions in tax payer money to increase the speed and competitiveness of the US networks,

So you will not see the "Next" increase come within the next 15 years i'm thinking.
 
Well light does travel pretty fast. It can circle the earth in about 130ms, so there is still room to improve on current fiber connections..
 
k1pp3r said:
In the US? You will not see the next speed jump for a long ass time, reason. . . . .Competition, ISP's have squandered billions in tax payer money to increase the speed and competitiveness of the US networks,

So you will not see the "Next" increase come within the next 15 years i'm thinking.
Click to expand...

The reason I heard is because the US started the internet and implemented it early. That means laying down the framework and backbone lines all around the country. At the time we did that, we probably did have the state of the art, but not any more. It's simply too costly to dig up the ground and lay more lines everywhere.

The other countries jumped on the bandwagon much later and consequently used the best stuff they could get their hands on at the time, plus they had massive government help.
 
prochobo said:
The reason I heard is because the US started the internet and implemented it early. That means laying down the framework and backbone lines all around the country. At the time we did that, we probably did have the state of the art, but not any more. It's simply too costly to dig up the ground and lay more lines everywhere.
Click to expand...

The $300 Billion that was squandered already would have been a great start, but noooooooooooo. Sorry, but a lot of countries have infrastructer hundreds of years old that are better than the US
 
Remember the infrastructure in the US is still old, and we cover great expanses. Add to that, the competition between different providers..including backbones.

Many countries that we envy for having fast connections..Korea, Japan, Sweden....they're small...they don't have to have backbones across a few thousand miles. And quite a few got leveled in relatively recent wars, so infrastructure was completely rebuilt and thus easier to upgrade.

In some of those small countries, where 99% of the population lives in a handful of very dense cities of super highrise habitats..it's easy to put out fast broadband to those little areas and then claim that the majority of your countries population has fast broadband.

Also, most of those countries...government subsidized ISPs.
 
There's little competition in the US, while at the same time the regulators are not pushing the companies they regulate to innovate. Until Verizon, AT&T, etc build out their networks and present a real competition to Comcast/Time Warner/etc there'll be little incentive to compete on quality. And even when that happens, a duopoly is not as good as a competitive market. We have an anti-competitive market. In such a market, the companies should be basically pushed by their regulators to provide better quality and to meet world standards...
 
Hey guys, I'm more interested in future technologies, not in economic or political reasons why utility companies suck (I know plenty about that already).

As for the physical limitations, I think eddie is correct:

eddie11c said:
Well light does travel pretty fast. It can circle the earth in about 130ms, so there is still room to improve on current fiber connections..
Click to expand...

That's the same number I calculated, using: (((40041.47)/2)/(299792.458))*2, which is the earth's mean circumference divided by 2, divided by the speed of light in kilometers, and multiplied by 2 because a ping has to go there and back.
 
InnocentNoobie said:
Hey guys, I'm more interested in future technologies, not in economic or political reasons why utility companies suck (I know plenty about that already).

As for the physical limitations, I think eddie is correct:
Click to expand...

Well for technology a few years ago there were some scientists that teleported a laserbeam 5m that was encoded with information.

Imagine what could be done with teleporting data as internet backbones.
 
eddie11c said:
Well light does travel pretty fast. It can circle the earth in about 130ms, so there is still room to improve on current fiber connections..
Click to expand...

After considering the refractive index of the actual optical fiber, it will take closer to 155ms to make a lap around the world. But that's not really what's important.

The biggest bottlenecks as it stands are the local loop latencies and the raw propagation delay across long distances.

As an example, if I want to send a packet from the east coast to Japan.

  • It takes anywhere from 5-100ms typically depending on the type of link you have to your ISP, just to reach the backbone pop for your region.
  • Then you're looking at a raw propagation delay of 55-60ms to blink a light here from the east coast backbone pop to Japan (depending on the refractive indices of the fiber along the way), and I'm leaving out serialization and other delays imposed by SP equipment hops along the path as well :)
  • Once the data crosses the ocean, you then have to wait for propagation delay from the SP pop to the home user endpoint over there which again could be 5-100ms.

See how it all starts to add up? In our tech utopia world, when everyone has FTTH with 5ms pings to the local backbone pop, you're still going to get between 70-100ms pings across the ocean in the best case scenario.

That's about as good as it's ever going to get, but I think we'll be able to cope with those delays. :cool:
 
eddie11c said:
Well light does travel pretty fast. It can circle the earth in about 130ms, so there is still room to improve on current fiber connections..
Click to expand...

How ever even high end telecom equipment still has attenuation to deal with, and I don't know if it's really possible to mitigate that any better then we already do. Also having a single fiber from Chicago to Tokyo is pretty much out of the question. With the Internets current design having multiple independent routers adds propagation delay and forwarding latency it adds the ability to tolerate medium failures. Even line-rate hardware routers add some degree of latency for each hop.

Unless we can start making our own little wormholes to mitigate the distance problem, or find some method of faster then light communication, I wouldn't expect latency to change much.
 
Before looking at the propogation across the ocean I think we have to look at our local backbones as Lightworker pointed out. The US is lagging badly in the internet race..
 
Well certain types of local loops just have poor latency to begin with. My DSL line at home takes on average 50ms to get to our remote equipment at work, which is less than two miles away. Same ISP too.

Cable isn't known for stellar latency either, but the FTTH and FTTC offerings like FiOS and U-Verse generally have pretty solid delay figures. As the "new" broadband market matures and grows those connections will be more widespread. :)
 
Well, if you're looking for better latencies for farther away places, all you can say would be a revolution is better point-to-point connections between you and the destination.

It's safe to say traffic comes over the Atlantic and lands in places like New York and hits a router there. ...But if it came over and went right to your city, certainly it would lose the ms caused by the hop in NY.

Not much of a revolution, just a more vast implementation of current technology I guess. But undersea cables are probably cost a pretty penny to build and maintain.


...or we could drill straight down and shoot lazorz through the earth. There's an idea.
 
Lightworker said:
That's about as good as it's ever going to get, but I think we'll be able to cope with those delays. :cool:
Click to expand...

Well, at least until we find a way to open inter-spatial tunnels that won't result in the destruction of our planet...

::looking down at my signature::


I'm curious what the actual limit in data bandwidth would be though. For a while, the limiting speed was how fast LEDs could be switched on and off, but I guess currently it's how fast a laser beam can be pulsed.

I suppose there has to be some physics limit to how fast the light could be pulsed, but not being a theoretical physicist, I can't say what that might be. The speed of light is a limiting factor when it comes to ping times, but what the actual data throughput limit would be an interesting discussion.

I'm thinking that there would be a certain point where, if the light burst was short enough, certain particles of scattered light thrown off into the optical fiber medium would follow a straighter path than other particles (bouncing back and forth along the fiber in a zig-zag) and as such would travel a slightly shorter distance. This would cause these light waves to reach the destination fractions (probably very small fractions) of a millisecond earlier than other light signals, and could cause an echo effect that might prevent the next bit from being received properly.

These are just my uninformed ramblings however, and hopefully can spark a more informed discussion. :)
 
Last edited:
You must log in or register to reply here.
Back
Top