for me, it's the 3 ms latency that i've been longing for....
*sighs*
i know right ?
i wonder if they have an FUP =P
Chup wrotei wonder if they have an FUP =P
Yeah, they probably throttle them from 523 Mbits to 500Mbits if they download more than two terabytes per day.
:D
The real question is: I wonder if they have any LAN in there?


Faster than 99% of the U.S... LoooL. Who wants a faster connection anyway ? (other than the pentagon)
I guess you guys are overlooking something essential here. It's true they have capped internet speed, but they are limited by the "Internet" speed. Google doesn't control the whole network.
It's like I am doing a ping test down the hall and saying I have 100MB/s internet.
Am I missing something?
Arithma, the server they are testing against is 50miles away from Google. But most probably, or certainly Google is connected to the Internet by the means of a fiber optic cable . That explains the very low latency which is only caused by the termination hardware on both sides . Actual signals down the line travel at the speed of light ! :)

Taybe lal google .
Google bought huge chunks of networks and IPs. There were rumors about Google entering the ISP market, but in never materialized. Now I see where those purchases went :-)
arithma wroteI guess you guys are overlooking something essential here. It's true they have capped internet speed, but they are limited by the "Internet" speed. Google doesn't control the whole network.
It's like I am doing a ping test down the hall and saying I have 100MB/s internet.
Am I missing something?
What you are saying is right. Having a 500 Mbit/s connection doesn't mean that people at Google can load all sites or download all movies and softwares on the internet at that speed... the speed will be limited by the other end, the web servers...

If for example lebgeeks.com is hosted on a web-server with a 100 Mbit/s Leased line, then Google at best conditions will not be able to load lebgeeks.com faster than 100 Mbit/s (best conditions mean that nobody in the whole world is loading lebgeeks.com at the same time, and nobody in the whole world is loading any other websites hosted at the same server...).

The same for the latency, it is related to how far the targeted servers are from Google headquarter, so if a guy at Google (California) wanted to play a FPS game and the game-server is located in Australia, then he will have a 300-500 ms latency... and as Jad said, latency is also related to the termination hardwares on both sides...
Evil speed thanks for sharing; I can't imagine NASA's connection!

By the way that's Apple's:
I think NASA's latency between its main HQ and one of its Satellites is 5 ms.
kareem_nasser wroteI think NASA's latency between its main HQ and one of its Satellites is 5 ms.
Impossible.
kareem_nasser wroteI think NASA's latency between its main HQ and one of its Satellites is 5 ms.
Lets do some simple childish calculations :

Not taking into consideration the lag of land and space equipment. we will see how much time it takes for the RF(Radio Frequency) signal itself to reach the satellite or the base station.

RF travels approximately as the speed of light which is = 300,000 km/s .

The distance to a medium earth orbit satellite is approximately 35,000 km .

35,000/300,000= 116.6 milliseconds for a one way trip !

Thats 233.3 milliseconds for a round trip in the BEST of cases !

The question is, did NASA break the speed of light for communication technology ?! I doubt :)
Guys, you have to start thinking of it this way:
The internet is a network. It's a graph. Much like our traffic in Lebanon. If people on the Saida/Beirut highway can reach 100Km/h it doesn't mean they can get anywhere around Lebanon that fast. The analogy is perfect, since you can think of your packets as cars moving around.
What they're measuring over there is internal traffic speed (50 mi for google is down the hallway for me, may be less).
J4D wrote
kareem_nasser wroteI think NASA's latency between its main HQ and one of its Satellites is 5 ms.
Lets do some simple childish calculations :

Not taking into consideration the lag of land and space equipment. we will see how much time it takes for the RF(Radio Frequency) signal itself to reach the satellite or the base station.

RF travels approximately as the speed of light which is = 300,000 km/s .

The distance to a medium earth orbit satellite is approximately 35,000 km .

35,000/300,000= 116.6 milliseconds for a one way trip !

Thats 233.3 milliseconds for a round trip in the BEST of cases !

The question is, did NASA break the speed of light for communication technology ?! I doubt :)
jad. this is too geeky to be true. you're damn good at this.
kareem, no offense my friend, but please make sure to post some correct values based on a scientific proof.

i also wonder what is the ultimate speed we can ever reach in terms of internet speed. because soon enough, the entire communication system we have will be outdated. and speeds will not be sufficient to fit our needs.
I think this is interesting http://en.wikipedia.org/wiki/COMMStellation

"COMMStellation will be situated at 1,000km from the surface of the Earth and will be connected to the Internet trunk via 20 ground stations located across the globe. Each of the microsatellites will have a data-transfer capacity of 12 gigabits per second"

"COMMStellation will provide 100% global coverage reaching areas of the world where equatorial orbiting constellations are ineffective, providing access to bandwidth to the entire population of the Earth"

Oh and about the latency: "The planned COMMStellation™, scheduled for launch in 2015, will orbit the earth at 1,000 km with a latency of approximately 7 ms" from http://en.wikipedia.org/wiki/Satellite_Internet_access
Georges wrotei also wonder what is the ultimate speed we can ever reach in terms of internet speed. because soon enough, the entire communication system we have will be outdated. and speeds will not be sufficient to fit our needs.
As far as i know, their is nothing faster than light !
We already use light as a means of communication with fiber optics.
But the limit now is with the semiconductors and the equipment found at each end.
You know of course that data travels in binary as zero's and one's.
So if i want to send the number 7 im going to send 0111 in a 4bit BCD notation.
Thats off-on-on-on if the binary was converted to light beams.
on the transmitting end the limit is the "transistor" that has to actually switch the light source on and off at blistering speed. On the receiving end is the "photo transistor" that has to detect when light is present or not on that certain line. their is a limit to how much the transistor can do the switching. Research in the semiconductor field is very rapid, and i dont think that we will face any problem in the future.

The good thing about using fiber optics is that you dont need to change the fiber-optic cable to get better speed.
you only change the equipment on both ends to something more capable ;)