What if networking technology improves to the point where sub-millisecond ping times become commonplace?
-
-
@El_Heffe said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Just move closer to the website, stupid.
-
@Lorne-Kates said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Just move closer to the website, stupid.
That's not necessary, we just build things smarter. The response to a PING is quite obvious, so your ISP could answer it without even going to website.
-
You mean like Google does with their AMP bullshit?
-
@Lorne-Kates said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Just move closer to the website, stupid.
Plenty of banks do exactly this, building a new data centre as close as possible to the transatlantic cable or whatever, because a millisecond improvement in latency can make enough of a difference to high frequency trading algorithms that it's worth the cost od the new data centre
-
@El_Heffe said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
The speed of light is unlikely to improve
Quantum entanglement networking!
-
@Jaloopa said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Plenty of banks do exactly this, building a new data centre as close as possible to the transatlantic cable or whatever, because a millisecond improvement in latency can make enough of a difference to high frequency trading algorithms that it's worth the cost od the new data centre
I believe they've gone to the point of co-hosting the HFT platforms with the exchange (and paying quite a bit for the privilege). Can't get much closer than that. The next logical step (and where things start to get truly nasty) is to deploy code on those systems that surreptitiously tries to slow down traffic to other HFT systems. At that point, you'll be in a race to running Core Wars except with trillions of dollars at stake…
-
@Jaloopa said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
high frequency trading algorithms
HFT is TRWTF, to be honest, and I'm not even an anti-1%er.
-
@El_Heffe said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
TRWTF on that page is the comments and/or the commenters. RC himself is on-target.
-
While milliseconds probably are quite sufficient, a 32bit integer and microseconds gives you an upper bound of a bit more than 1h. Ping times approaching 1h are probably fairly well approximated with E_NO_CONNECTION (preferably a bit before the 1h elapses). Besides, 64-bit integers do exist these days. So, what's the argument against µs?
What you'd show the user is a different question (the post makes it sound like an API choice), but ping et al. solved that problem a long time ago.
-
@Steve_The_Cynic said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
HFT is TRWTF, to be honest, and I'm not even an anti-1%er.
The number of times a minor bug in a HFT algo has wiped millions off a stock market, I can't disagree
-
@cvi said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Ping times approaching 1h are
probably fairly well approximated with E_NO_CONNECTIONto be expected in interplanetary communicationsNo playing CoD with Martian explorers in their off-duty times!
-
@fbmac said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
The response to a PING is quite obvious, so your ISP could answer it without even going to website.
Or your NIC could just respond immediately. There's your sub-millisecond ping.
-
@Bort said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@fbmac said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
The response to a PING is quite obvious, so your ISP could answer it without even going to website.
Or your NIC could just respond immediately. There's your sub-millisecond ping.
Sometimes graphics drivers do shady things like that to improve benchmark scores. Hopefully AMD and NVIDIA never get into high-performance NICs.
-
@cvi I agree. What if someone wants to see the difference between 2ms and 2.5 ms?
-
@dkf said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@cvi said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Ping times approaching 1h are
probably fairly well approximated with E_NO_CONNECTIONto be expected in interplanetary communicationsAnd as if timezones weren't hard enough, you now have to deal with time dilation.
-
@anonymous234 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@dkf said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@cvi said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Ping times approaching 1h are
probably fairly well approximated with E_NO_CONNECTIONto be expected in interplanetary communicationsAnd as if timezones weren't hard enough, you now have to deal with time dilation.
meh. if you're just trying to get a data link between say earth and mars you can actually mostly ignore time dilation, the timing tolerances built into the communications protocols will more than compensate for the nanoseconds of drift between Eath and Mars.
if you're going to be building an accurate solar system positioning system on the other hand..... yeah you need to take tidi, general relativity AND special relativity into account.
-
@anonymous234 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@cvi I agree. What if someone wants to see the difference between 2ms and 2.5 ms?
People who want that are most likely using specialized hardware that allows for that. Our NICs and associated software measure timing in the microsecond range, but obviously it's specialized hardware for specialized purposes. Though apparently we have plans to create an NDIS driver for them so Windows can use it like any other NIC...that will be an interesting project.
-
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Though apparently we have plans to create an NDIS driver for them so Windows can use it like any other NIC...that will be an interesting project.
:static: Attention, this is the captain. We're currently first in line for takeoff, but we're waiting for Windows Updates to complete and reboot before the plane will be able to fly. I'll get back to you when the situation has updated. :static:
-
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Though apparently we have plans to create an NDIS driver for them so Windows can use it like any other NIC...that will be an interesting project.
:static: Attention, this is the captain. We're currently first in line for takeoff, but we're waiting for Windows Updates to complete and reboot before the plane will be able to fly. I'll get back to you when the situation has updated. :static:
Thank God there is no real-time1 version of Windows 10. Planes would literally be falling out of the sky because Windows Updates are more important than maintaining level flight.
1I really wanted to add the term "flight-certified" but gut instinct tells me a real-time Win10 would find its way onto a flight computer whether it was certified or not.
-
@Bort said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Or your NIC could just respond immediately. There's your sub-millisecond ping.
I already get super-low latency when I ping
127.0.0.1
…
-
@dkf said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@Bort said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Or your NIC could just respond immediately. There's your sub-millisecond ping.
I already get super-low latency when I ping
127.0.0.1
…What we need next is Reverse Cloud Computing. If you could run everything off your local PC your latency would be great!
-
@Scarlet_Manuka spotting!
-
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Reverse Cloud Computing
?gnitupmoC duolC
-
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
If you could run everything off your local PC your latency would be
greatsmall!that for you
-
@fbmac I walked right into that one.
-
@dkf said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@cvi said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Ping times approaching 1h are
probably fairly well approximated with E_NO_CONNECTIONto be expected in interplanetary communicationsNo playing CoD with Martian explorers in their off-duty times!
I'm not going to do the math, but I think even the worst case (earth -> earth L4/5 -> mars) with earth/mars on opposite sides of the sun can be done in under half an hour? Planetary transfer times are insignificant in that case, of course.
I'm also pretty sure I've seen fractional millisecond ping times between my linux PC and my router.
-
@dkf said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@cvi said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Ping times approaching 1h are
probably fairly well approximated with E_NO_CONNECTIONto be expected in interplanetary communicationsNo playing CoD with Martian explorers in their off-duty times!
How about a nice game of
chessinterplanetary accelerated-mass bombardment?
-
@PleegWat said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
I'm not going to do the math, but I think even the worst case (earth -> earth L4/5 -> mars) with earth/mars on opposite sides of the sun can be done in under half an hour? Planetary transfer times are insignificant in that case, of course.
Still better than MilwaukeePC /cc @ben_lubar
-
@PleegWat said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
I'm also pretty sure I've seen fractional millisecond ping times between my linux PC and my router.
Yep.
$ ping 192.168.1.1 PING 192.168.1.1 (192.168.1.1) 56(84) bytes of data. 64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=0.462 ms 64 bytes from 192.168.1.1: icmp_seq=2 ttl=64 time=0.445 ms ^C --- 192.168.1.1 ping statistics --- 2 packets transmitted, 2 received, 0% packet loss, time 999ms rtt min/avg/max/mdev = 0.445/0.453/0.462/0.022 ms
-
Followup, related only by the 'speed of light reference' and nothing else:
The speed of light configuration is totally unsupported. Since all the power management safeguards were disabled, the thermal sensor that said "Whoa, I'm running really hot! Could you drop to low power for a while?" was being ignored. A common side effect of the speed of light configuration was random crashes due to overheating.
-
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Though apparently we have plans to create an NDIS driver for them so Windows can use it like any other NIC...that will be an interesting project.
:static: Attention, this is the captain. We're currently first in line for takeoff, but we're waiting for Windows Updates to complete and reboot before the plane will be able to fly. I'll get back to you when the situation has updated. :static:
Thank God there is no real-time1 version of Windows 10. Planes would literally be falling out of the sky because Windows Updates are more important than maintaining level flight.
1I really wanted to add the term "flight-certified" but gut instinct tells me a real-time Win10 would find its way onto a flight computer whether it was certified or not.
Thankfully, it's unlikely that the familiar FMC will ever be ported to a Windows edition.
And one for the bad ideas thread: fly-by-cloud-wire. It'll save a couple pounds of computing systems, after all. Just send all of the pilot's control commands to the cloud, process it, and send it back!
-
@Sumireko said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
And one for the bad ideas thread: fly-by-cloud-wire. It'll save a couple pounds of computing systems, after all. Just send all of the pilot's control commands to the cloud, process it, and send it back!
That's not the bad-idea-thread, that's the terror-inducing--thread. Or the approved-by-terrorist thread...
-
@Sumireko said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
And one for the bad ideas thread: fly-by-cloud-wire. It'll save a couple pounds of computing systems, after all. Just send all of the pilot's control commands to the cloud, process it, and send it back!
But how else will Google force you to watch 30-second video ads before you're allowed to deploy the landing gear, or get you up-to-date online help?
-
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
But how else will Google force you to watch 30-second video ads before you're allowed to deploy the landing gear, or get you up-to-date online help?
You've only got 8 hours to fly that plane around the world before it shuts down to install a critical update to Candy Crush.
-
@dkf said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@Jaloopa said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Plenty of banks do exactly this, building a new data centre as close as possible to the transatlantic cable or whatever, because a millisecond improvement in latency can make enough of a difference to high frequency trading algorithms that it's worth the cost od the new data centre
I believe they've gone to the point of co-hosting the HFT platforms with the exchange (and paying quite a bit for the privilege). Can't get much closer than that. The next logical step (and where things start to get truly nasty) is to deploy code on those systems that surreptitiously tries to slow down traffic to other HFT systems. At that point, you'll be in a race to running Core Wars except with trillions of dollars at stake…
Pretty much (minus the Neuromancer-esque Core Wars (which would be awesome until the inevitable collapse)).
https://www.amazon.com/Flash-Boys-Wall-Street-Revolt/dp/0393351599
-
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Quantum entanglement networking!
spooky bugs at a distance
-
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
@PleegWat said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
I'm also pretty sure I've seen fractional millisecond ping times between my linux PC and my router.
Yep.
$ ping 192.168.1.1 PING 192.168.1.1 (192.168.1.1) 56(84) bytes of data. 64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=0.462 ms 64 bytes from 192.168.1.1: icmp_seq=2 ttl=64 time=0.445 ms ^C --- 192.168.1.1 ping statistics --- 2 packets transmitted, 2 received, 0% packet loss, time 999ms rtt min/avg/max/mdev = 0.445/0.453/0.462/0.022 ms
What kind of shit-grade router are you running there? I'm seeing rtt min/avg/max/mdev = 0.213/0.222/0.236/0.008 ms, and that's from a router two switch hops away from my PC.
-
@flabdablet said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
What kind of shit-grade router are you running there?
It's whatever Verizon gave me.
@flabdablet said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
I'm seeing rtt min/avg/max/mdev = 0.213/0.222/0.236/0.008 ms, and that's from a router two switch hops away from my PC.
It's probably because you're using metric milliseconds.
-
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
metric milliseconds
Wouldn't an imperial second be subdivided in 60 tertians or something?
-
@mott555 said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Sometimes graphics drivers do shader-y things
-
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
you're using metric milliseconds
Nah, it will be our superior mains voltage. You can't expect your packets to fly round your network with anything approaching enthusiasm if their only motivation is a feeble 110 volts.
-
@flabdablet said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
What kind of shit-grade router are you running there?
Apparently something somewhere between what you have and he has:
64 bytes from 192.168.1.1: icmp_seq=0 ttl=64 time=0.433 ms 64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=0.358 ms 64 bytes from 192.168.1.1: icmp_seq=2 ttl=64 time=0.380 ms 64 bytes from 192.168.1.1: icmp_seq=3 ttl=64 time=0.369 ms 64 bytes from 192.168.1.1: icmp_seq=4 ttl=64 time=0.375 ms 64 bytes from 192.168.1.1: icmp_seq=5 ttl=64 time=0.384 ms 64 bytes from 192.168.1.1: icmp_seq=6 ttl=64 time=0.417 ms ^C --- 192.168.1.1 ping statistics --- 7 packets transmitted, 7 packets received, 0.0% packet loss round-trip min/avg/max/stddev = 0.358/0.388/0.433/0.025 ms
-
@boomzilla said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
It's whatever Verizon gave me.
Well that's why... ISP routers are categorically awful
-
@flabdablet said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Nah, it will be our superior mains voltage.
Yeah, our mains voltage is YUUUGE!
-
@dkf US mains voltage a disaster. SAD!
-
@El_Heffe said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
It's a "what if". You have to assume whatever is needed for that to happen to be solved to play along.
-
@sloosecannon said in What if networking technology improves to the point where sub-millisecond ping times become commonplace?:
Well that's why... ISP routers are categorically awful
How could I tell the difference?
-
@boomzilla Try running a torrent session with 1000 peers. Most ISP-provided routers will sulk very badly if required to maintain more than about a hundred simultaneous NAT mappings.
-
@flabdablet Yeah, I've never done much torrenting.