blackened23
Diamond Member
- Jul 26, 2011
- 8,548
- 2
- 0
i think that, diablo 3 shows us that, online games just don't work
(on skype) Alright guys, watch my dps on skeleton king! (buffering...)
(buffering...{)
i think that, diablo 3 shows us that, online games just don't work
I realize DVD is old algorithm but its fast to encode and decode, remember any compressing they do of the output file(H.264 or whatever they go with) is going to add latency which is already going to be high with this type of setup, they would have to go with a very fast algorithm that doesnt take much CPU/GPU time to encode/decode for best performance and thats going to result in larger files to stream.
100-200 ms is pretty normal. I have seen much worse... A really good broadband can get much lower though
0.65c, aka 200km/ms. Is the speed of light in a fiber optic cable.
the US is 3000 miles across (east to west)
3000 miles = 4 828 032 meters
It takes a mere 24.14016ms for light to cross that distance in a fiber optic cable.. And usually the servers are closer. The vast majority of your ping is coming from switching. (the fact the cables are not going in a straight line adds a bit too)
It will be obscene...
yea but DVDs are encoded in MPEG2, an algorithm from 1996.
Still, its going to be an obscene amount of data that will make ISPs throw a hissy fit.
False.
The vast majority of lantency comes from a) protocol lantency and b) distance.
You could read up here as I have covered this before:
http://forums.anandtech.com/showthread.php?t=2085280&highlight=shdsl&page=2
Post #38
(Did you forget that thread already? )
That you think 200ms ping is normal, tells me that either you are not informed...or the ISP's in the US are very abysmal
I didn't forget, I was ignoring you. You are being rude about it so it doesn't deserve an answer. Also you seem to be skimming my posts before doing so and then copy pasting the exact same reply post
No it dosn't
Electricity moves at ~66% the speed of c in a shielded copper cable.
But we don't don't use copper in the internetnet backbone, we use fiber.
But there the same rules apply, light only moves at ~66% the speed of c in a fiber cable.
Simple physics FYI.
The only place you can reach c is in a vacuum.
You need to redo your math.
Oh, BTW you need to learn something about networking too:
Code:Tracing route to [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL] [208.65.201.105] over a maximum of 30 hops: 1 <1 ms * <1 ms xxx.xxx.xxx [xxx.xxx.xxx.xxx] 2 10 ms <1 ms <1 ms ve8.cosw1.hoer.dk.ip.fullrate.dk [90.185.3.25] 3 1 ms 1 ms 1 ms te2-2.boanxc7.dk.ip.tdc.net [195.215.109.229] 4 31 ms 16 ms 16 ms xe-0-3-0.cr1.lhr1.uk.nlayer.net [195.66.224.37] 5 95 ms 107 ms 94 ms xe-7-0-0.cr1.nyc3.us.nlayer.net [69.22.142.30] 6 100 ms 100 ms 99 ms xe-2-0-0.cr1.iad1.us.nlayer.net [69.22.142.92] 7 100 ms 100 ms 100 ms ae1-40g.ar1.iad1.us.nlayer.net [69.31.31.178] 8 100 ms 100 ms 100 ms r1.vadc1.eicomm.net [69.31.30.218] 9 100 ms 100 ms 100 ms 208.65.201.105 Trace complete.
Notice how switching/routing adds no real measurable lantency, but distance does?
That because the biggest fator of lantency in world wide communication is distance, not the number op hops.
Network Technician (with an understanding of physics) for an ISP here.
Just F.Y.I.
A ping is too and from.
A sends a packet to B
B recieves packet and sends reply back.
So you need to double the distances/time.
And you forget that I am sitting on a 1G fiberlink directy into out backbone, so I don't have the intial lantency on various xDSL technologies.
SHDSL = 1-3ms latency from IAD til DSLAM
ADSL = 8-15ms lantecy from IAD to DSLAM
ADSL2+ = 20-35 ms lantecy from IAD to DSLAM
I know, have a KillerNic myself, due to the fact that I can compile linux apps to run directly on my NIC.
Onboard NIC's are crap...especially if you put a heavy load on your connection, or run many sessions at the same time.
I will dissect it here:
Code:1 <1 ms * <1 ms xxx.xxx.xxx [xxx.xxx.xxx.xxx] [B]My connection[/B] 2 10 ms <1 ms <1 ms ve8.cosw1.hoer.dk.ip.fullrate.dk [90.185.3.25] [B]Our coreswitch...notice the lantency is the same, even if has been routed.[/B] 3 1 ms 1 ms 1 ms te2-2.boanxc7.dk.ip.tdc.net [195.215.109.229] [B]Another ISP, but in the same country...notice the same lantency, we have routed but travled not far[/B] 4 31 ms 16 ms 16 ms xe-0-3-0.cr1.lhr1.uk.nlayer.net [195.66.224.37] [B]Notice the ping goes up...moving Denmark to UK[/B] 5 95 ms 107 ms 94 ms xe-7-0-0.cr1.nyc3.us.nlayer.net [69.22.142.30] [B]Notice the pings go up...moving from the UK to the US[/B] 6 100 ms 100 ms 99 ms xe-2-0-0.cr1.iad1.us.nlayer.net [69.22.142.92] [B]Moving in the US[/B] 7 100 ms 100 ms 100 ms ae1-40g.ar1.iad1.us.nlayer.net [69.31.31.178] [B]In the US...but no added lantency, simple routing[/B] 8 100 ms 100 ms 100 ms r1.vadc1.eicomm.net [69.31.30.218] [B]In the US...but no added lantency, simple routing[/B] 9 100 ms 100 ms 100 ms 208.65.201.105 [B]In the US...but no added lantency, simple routing and at destination[/B]
The ping show pings form different locations(hops) on the router from my IP to anandtech.com, but my ping should be ~100 ms til anandtech.com from that traceroute:
Code:Pinging [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL] [208.65.201.105] with 32 bytes of data: Reply from 208.65.201.105: bytes=32 time=100ms TTL=59 Reply from 208.65.201.105: bytes=32 time=100ms TTL=59 Reply from 208.65.201.105: bytes=32 time=100ms TTL=59 Reply from 208.65.201.105: bytes=32 time=100ms TTL=59 Ping statistics for 208.65.201.105: Packets: Sent = 4, Received = 4, Lost = 0 (0% loss), Approximate round trip times in milli-seconds: Minimum = 100ms, Maximum = 100ms, Average = 100ms
And so it is.
You to take into account that a core-router has pings and trace as VERY low priority.
It handlles the packte traffic and when it has ilde time it repplies on ping/traces, that is why the number fluctuates.
And why you look at the first number (the "norm")
This is from a customers router (running ADSL2+)
Code:fullrate> ip traceroute [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL] Resolving [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL]... traceroute to [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL] (208.65.201.105) 30 hops max, 40 byte packet 1:xxx.xxx.xxx.xxx (50 ms) (20 ms) (20 ms) 2:90.185.4.217 (v243-cosw1-ar.fullrate.dk) (20 ms) (20 ms) (20 ms) 3:195.215.109.205 (te2-2.arcnxc7.dk.ip.tdc.net) (20 ms) (20 ms) (20 ms) 4:83.88.13.5 (pos6-0-0.ldn2nxg2.uk.ip.tdc.net) (50 ms) (40 ms) (50 ms) 5:195.66.226.37 (xe-3-0-0.cr1.lhr1.uk.nlayer.net) (40 ms) (40 ms) (50 ms) 6:69.22.142.9 (xe-2-2-0.cr1.nyc3.us.nlayer.net) (120 ms) (130 ms) (120 ms) 7:69.22.142.92 (xe-2-0-0.cr1.iad1.us.nlayer.net) (130 ms) (130 ms) (130 ms) 8:69.31.31.178 (ae1-40g.ar1.iad1.us.nlayer.net) (130 ms) (130 ms) (130 ms) *snip*
Notice the 20 ms at the first hop?
That was the xDSL lantency from IAD(router) to DSLAM
But notice that even if we route from hop 2 to 3, the lantency don't go up, because the geopgrahical distance is miniscule.
The biggest step in lantency is then:
From IAD to DSLAM (hop 1) 20ms added
From DK to UK (hop 4) 30ms added
From UK to US (hop 6) 80ms added
At total of 130ms added from geographical distance.
Now if we ping from the same router, we get this:
Code:fullrate> ip ping [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL] Resolving [URL="http://www.anandtech.com/"][COLOR=#0066cc]www.anandtech.com[/COLOR][/URL]... 208.65.201.105 sent rcvd rate rtt avg mdev max min 1 1 100 130 130 0 130 130 2 2 100 120 129 3 130 120 3 3 100 120 128 5 130 120
The lantecy is added via distance...not routing.
Hope that helped.
You're ignoring the facts about bandwidth needed for this to run properly, the fact that most ppl have ADSL internet (latency), slower than required speed, they quality of the encoded video and the cost of encoding it at 60 fps at decent quality.
Problems, problems everywhere.
I was referring to Lonbjerg. Sorry I didn't check your post before posting.
You seem to be under the mistaken impression that sending a packet from US, to denmark, to UK, back to the US means that the issue is one of "distance". Its not, its a switching problem where packets are routed incorrectly.
You further claimed that the majority of ping is from the network stack... yet the example YOU PROVIDED showed it taking LESS than 1ms to bounce the local router proving that the network stack took under 1ms.
In comparison, it took 10 ms to bounce it from local ISP (which is way way closer then 2 million meters = 1,242.74238 miles.)
Of those 10 ms, it was neither the network stack nor the distance that accounted for the majority. That only leaves the ISPs switches/routers.
Then the ISP shitty system makes a critical error and tosses the connection across the ocean. That adds a serious amount of lag. Distance is PART of the issue but that is distance that shouldn't have been traveled. If I drive my car in circles around the destination then the reason I am using up a lot of fuel is not the distance to my target, its the fact I don't know how to read a map.
I offered no counter earlier because you are extremely rude and I have better things to do with my time then raise my blood pressure at someone who doesn't know basic civility.
You're ignoring the facts about bandwidth needed for this to run properly, the fact that most ppl have ADSL internet (latency), slower than required speed, they quality of the encoded video and the cost of encoding it at 60 fps at decent quality.
Problems, problems everywhere.
Cloud gaming will be fine for old school adventure game where latency is not much of an issue, but on the other hand any modern smart phone can run old school adventure game just fine natively.
Imagine playing some latest pc/console games on a smart phone on the go... only realize you only have 2GB of data per month.
What speed (Mbit/s)is required?
What lantency (ping) is required?
Those are the part that is in my field of work and it seems to me a lot of posters inhere are cluelss about networking and post rubbish, even when informaed of the facts...and that they are wrong...they go "not see, not hear, only post false rubbish.
Come back when you have CCNA or plus...sorry to say it...but you are babbling about stuff your cleary know nothing about...it's painfull to watch.
For an h264 1080p 3 hours action movie like Watchmen you need about 20 GB.
But you can't have most of the encoder's prediction features (frame refs, temporal, MB-Tree, etc) since it can't check future frames and that fact would skyrocket the bitrate needed to way above the max specs of bluray standard.
That sole thing is leaving out any connection with less than 50 Mbps. But you said that anything below 60 fps is unacceptable (as it is in consoles) so that bitrate will be even higher without any prediction to lower it.
So it leaves lossless or semi-lossless with about the same or even higher bitrate.
As for latency now I'm sitting on a DOSCIS 3.0 connection with 10 ms to my country (~50 ms outside) but when I was using ADSL it was 50 ms (~100 ms outside). I bet this service won't place servers every 500 KM.
Aka I am right and you have no counter argument
I on the other hand (with others) have proven you to be false in this thread:
http://forums.anandtech.com/showthread.php?p=30046397#post30046397
You seem to be under the mistaken impression that sending a packet from US, to denmark, to UK, back to the US means that the issue is one of "distance". Its not, its a switching problem where packets are routed incorrectly.
You further claimed that the majority of ping is from the network stack... yet the example YOU PROVIDED showed it taking LESS than 1ms to bounce the local router proving that the network stack took under 1ms.
In comparison, it took 10 ms to bounce it from local ISP (which is way way closer then 2 million meters = 1,242.74238 miles.)
Of those 10 ms, it was neither the network stack nor the distance that accounted for the majority. That only leaves the ISPs switches/routers.
Then the ISP shitty system makes a critical error and tosses the connection across the ocean. That adds a serious amount of lag. Distance is PART of the issue but that is distance that shouldn't have been traveled. If I drive my car in circles around the destination then the reason I am using up a lot of fuel is not the distance to my target, its the fact I don't know how to read a map.
I offered no counter earlier because you are extremely rude and I have better things to do with my time then raise my blood pressure at someone who doesn't know basic civility.
I don't care about networking. I'm telling you the bitrate needed.
1080p * 60fps * live enconding without buffering and prediction * likely action on screen = shitload of bitrate that most of internet connections won't be able to handle.
Add metered connections as a trend today and you have a service doomed before even starting.
Keep raging and repeating yourself all you want. This service with better image quality than current consoles isn't feasible given the current internet connections.
I don't care about networking. I'm telling you the bitrate needed.
1080p * 60fps * live enconding without buffering and prediction * likely action on screen = shitload of bitrate that most of internet connections won't be able to handle.
Add metered connections as a trend today and you have a service doomed before even starting.
Keep raging and repeating yourself all you want. This service with better image quality than current consoles isn't feasible given the current internet connections.
I don't care about networking. I'm telling you the bitrate needed.
1080p * 60fps * live enconding without buffering and prediction * likely action on screen = shitload of bitrate that most of internet connections won't be able to handle.
Add metered connections as a trend today and you have a service doomed before even starting.
Keep raging and repeating yourself all you want. This service with better image quality than current consoles isn't feasible given the current internet connections.
No you didn't. You pointed out that I was not accounting for fiber optic being ~2/3rd the speed of light. Which while I didn't, its no longer the case (simple oversight on my part back then)... And does not in any way change the argument.
*SNIP*
100-200 ms is pretty normal. I have seen much worse... A really good broadband can get much lower though
0.65c, aka 200km/ms. Is the speed of light in a fiber optic cable.
the US is 3000 miles across (east to west)
3000 miles = 4 828 032 meters
It takes a mere 24.14016ms for light to cross that distance in a fiber optic cable.. And usually the servers are closer. The vast majority of your ping is coming from switching. (the fact the cables are not going in a straight line adds a bit too)
I will note debate on a foundation of lies:
The vast majority of your ping is coming from switching.
ARGUMENTS NOT BASED IN NETWORKING REALITY.
You seem to be under the mistaken impression that sending a packet from US, to denmark, to UK, back to the US means that the issue is one of "distance". Its not, its a switching problem where packets are routed incorrectly.
You further claimed that the majority of ping is from the network stack... yet the example YOU PROVIDED showed it taking LESS than 1ms to bounce the local router proving that the network stack took under 1ms.
In comparison, it took 10 ms to bounce it from local ISP (which is way way closer then 2 million meters = 1,242.74238 miles.)
Of those 10 ms, it was neither the network stack nor the distance that accounted for the majority. That only leaves the ISPs switches/routers.
Then the ISP shitty system makes a critical error and tosses the connection across the ocean. That adds a serious amount of lag. Distance is PART of the issue but that is distance that shouldn't have been traveled. If I drive my car in circles around the destination then the reason I am using up a lot of fuel is not the distance to my target, its the fact I don't know how to read a map.
I offered no counter earlier because you are extremely rude and I have better things to do with my time then raise my blood pressure at someone who doesn't know basic civility.
Again, what console (and game) runs 1080p @ 60 FPS.
It a really simple question.