I got 50/50 Internet and my bottlenecks tend to be on my side. Thank your ISP for giving your speed test high priority, but throttling everything else.
My low-end($600 band-new 6 years ago) Dell desktop can transfer a sustained 112MBytes/s while using under 0.1% cpu over SMB on my home network with the integrated NIC. Yes, the motherboard is 6 years old, but the HD is an SSD, so I can get those transfer rates over 1Gb.
Lets see what features an integrated NIC has. Adaptive Interrupt Moderation, Large Send Offload v2(IPv4/IPv6), and TCP/UDP(IPv4/IPv6) offloading.
Do we need a quad core cpu in a smartphone? Here we are, selling them like hot-cakes and it's growing the economy.
If you haven't realized, people make new services once the technology becomes available everywhere and cheap to purchase. Now if you don't think we shouldn't advance technology, why don't you go get an 8086 and pretend it does everything you "need" it to do.. oh.. wait... our entire economy is driven by luxury and technology, along with society. Lets just get rid of society because we don't "need" it.
You people said the same thing 10 years ago about broadband. "What do people need more than 56k for? Web pages load fast enough". Yet, here we are, 10 years later streaming HD videos as the norm.
"Right around the corner" as HP and Hynix have already stated that they are currently in the process of replacing a large portion of their production for memresistors for later this year. They were planning to do memresistors early this year, but they realized that memresistors are so much better than FLASH and memory, that they're going to eat into their own markets, so the postponed for Q4 '13.
Those tests are easy is mess with. ISPs already use traffic shaping to mess with speedtests to make them look better. Not to mention that those tests waited until the user was not using the Internet when doing the tests, meaning most tests would be done during off-peak hours.
Intel stated the 10Gb is trivial to implement using 22nm tech and expect to be integrated into low-end chipsets when it happens. The problem is 22nm is current tech and CPUs get priority. We need to wait a bit before 22nm trickles down to chipsets.
Doubling Internet speeds gives a fairly consistent 0.7% increase in GDP. Now go from 30Mb to 1Gb and a nation wide fiber roll-out will pay itself back in less than 1 year, assuming the average.
wiki: In measurements made between January and June 2011, the United States ranked 26th globally in terms of the speed of its broadband Internet connections, with an average measured speed of 4.93 Mbit/s
And I call BS on people having usable 30Mb connections. How many people actually get their rated speeds most of the time when they "need" it? I know there are boards full of people who complain that they nothing but unstable Internet, but they have no other options.
All of the theory that I learned in class-room has been valid since the 60s. The same basic patterns keep showing up, just with different names and market buzz.
SQL!! awesome!.. oh, you mean set theory
NoSQL!!! awesome!!. ohh, you mean hash tables?
More marketing hype about new tech!!! ohh, that again? What is it's name thing time?
Nothing is new, just new ways of implementing the old. Not to detract from a better way of using the old idea.
I can imagine that in the IPv6 age ISPs might have similar business reasons to limit the addresses they make available
They have to make available at least a/64 as the IPv6 standard requires it. Not handing out 2^64 address to each customer would cause Windows/Linux/MacOSX to break.
Except that the privacy addresses change every 15 minutes-ish and each application could have a different IP address for every request if wanted. DHCP on IPv4 cannot change your IP address without breaking your connections, IPv6 does not do that because it is generally understood that you will have lots of IP addresses.
You're not supposed to subnet past the/64, so you'll still need more than a/64 to properly segment your LAN. While the number of routable IP addresses will increase, the number of routes will go down, which is all the router cares about.
Assuming even 5000 people have Internet through my ISP, that is a bit over 10 chassis, each with 4 100Gb uplinks. They don't need to use all of them, but I drool at the thought of a switch or router with 40x100Gb ports. I understand that my ISP is not going to have a 4Tb Internet connection, but the internal network has the potential of a lot of inter-chassis bandwidth.
Many DOCSIS2.0 cable modems and all DOCSIS3.0. The top few cable companies, which represent almost 70% of the USA's broadband base, are nearly done with their DOCSIS 3.0 rollout. Rule of thumb, if it supports more than 30Mb, it is IPv6.
Good news, the Internet backbone has been IPv6 for over 10 years now and Cable modem and DSL hardware in the past 5-8 years have all supported IPv6 natively also.
As for "most software". Well, the most commonly used software is the web-browser, which has been IPv6 for a while now also. Most people purchase new hardware on a 5-8 year cycle and nearly all networking hardware in the past few years has been IPv6.
All 4g smartphones are IPv6, it is a requirement for 4g.
Paint on a tank, it makes the tank more resistant to attack. Think of all of the energy dissipated when the paint flakes off from a rocket hitting the armor!
This has been impressively ameliorated by the use of NAT, which shares numerous intenral and protected IP addresses behind a single or pair of public addresses
Yet another person who doesn't understand NAT. How do these people get jobs in networking?!
If we consumed 1,000,000/64s per second, it would take about 585,000 years to deplete IPv6. That is 1.8447x10^25 ip address per second for 580 millennia.
My ISP didn't give me a router, they gave me Gigabit Ethernet Switch+Bridge+Fiber Transceiver. It's like a big LAN, where each customer gets their own vLAN with a gateway which is the chassis. Except, instead of a regular 48 port switch, it's a 480 port Layer3 IPv4/IPv6 chassis with a 2Tb backplane and 4x100Gb uplinks.
My question is what do they plug those uplink ports into. Each chassis can only support 480 people and there are thousands. What does an aggregate switch for 100Gb ports look like? And what router accepts 100Gb ports?
And it may actually take a 100 days if you actually have to bother with testing it in all sort of scenarios, time zones, browsers en daylight savings time etc.
Send UTC and let the browser format and adjust to the local. VERY easy, almost no testing.
Slashdot Mirror
I got 50/50 Internet and my bottlenecks tend to be on my side. Thank your ISP for giving your speed test high priority, but throttling everything else.
You should see the benchmarks showing a FreeBSD box pushing 10Gb/s of 64byte packets with a single core x86 900mhz.
1080p BluRay is about 40Mb, which is 5MB/s. I have no idea how you come to the conclusion that 2160p can be done for 5MB/s without dropping quality.
My low-end($600 band-new 6 years ago) Dell desktop can transfer a sustained 112MBytes/s while using under 0.1% cpu over SMB on my home network with the integrated NIC. Yes, the motherboard is 6 years old, but the HD is an SSD, so I can get those transfer rates over 1Gb.
Lets see what features an integrated NIC has. Adaptive Interrupt Moderation, Large Send Offload v2(IPv4/IPv6), and TCP/UDP(IPv4/IPv6) offloading.
Do we need a quad core cpu in a smartphone? Here we are, selling them like hot-cakes and it's growing the economy.
If you haven't realized, people make new services once the technology becomes available everywhere and cheap to purchase. Now if you don't think we shouldn't advance technology, why don't you go get an 8086 and pretend it does everything you "need" it to do.. oh.. wait... our entire economy is driven by luxury and technology, along with society. Lets just get rid of society because we don't "need" it.
You people said the same thing 10 years ago about broadband. "What do people need more than 56k for? Web pages load fast enough". Yet, here we are, 10 years later streaming HD videos as the norm.
"Right around the corner" as HP and Hynix have already stated that they are currently in the process of replacing a large portion of their production for memresistors for later this year. They were planning to do memresistors early this year, but they realized that memresistors are so much better than FLASH and memory, that they're going to eat into their own markets, so the postponed for Q4 '13.
They will be selling both DDR3 and SSDs by 2015.
Those tests are easy is mess with. ISPs already use traffic shaping to mess with speedtests to make them look better. Not to mention that those tests waited until the user was not using the Internet when doing the tests, meaning most tests would be done during off-peak hours.
Intel stated the 10Gb is trivial to implement using 22nm tech and expect to be integrated into low-end chipsets when it happens. The problem is 22nm is current tech and CPUs get priority. We need to wait a bit before 22nm trickles down to chipsets.
And memresistors are right around the corner and can run at main memory speeds. How long until 4GB/s cheap SSDs?
Doubling Internet speeds gives a fairly consistent 0.7% increase in GDP. Now go from 30Mb to 1Gb and a nation wide fiber roll-out will pay itself back in less than 1 year, assuming the average.
wiki: In measurements made between January and June 2011, the United States ranked 26th globally in terms of the speed of its broadband Internet connections, with an average measured speed of 4.93 Mbit/s
And I call BS on people having usable 30Mb connections. How many people actually get their rated speeds most of the time when they "need" it? I know there are boards full of people who complain that they nothing but unstable Internet, but they have no other options.
All of the theory that I learned in class-room has been valid since the 60s. The same basic patterns keep showing up, just with different names and market buzz.
SQL!! awesome!.. oh, you mean set theory
NoSQL!!! awesome!!. ohh, you mean hash tables?
More marketing hype about new tech!!! ohh, that again? What is it's name thing time?
Nothing is new, just new ways of implementing the old. Not to detract from a better way of using the old idea.
I can imagine that in the IPv6 age ISPs might have similar business reasons to limit the addresses they make available
They have to make available at least a /64 as the IPv6 standard requires it. Not handing out 2^64 address to each customer would cause Windows/Linux/MacOSX to break.
Except that the privacy addresses change every 15 minutes-ish and each application could have a different IP address for every request if wanted. DHCP on IPv4 cannot change your IP address without breaking your connections, IPv6 does not do that because it is generally understood that you will have lots of IP addresses.
You're not supposed to subnet past the /64, so you'll still need more than a /64 to properly segment your LAN. While the number of routable IP addresses will increase, the number of routes will go down, which is all the router cares about.
Assuming even 5000 people have Internet through my ISP, that is a bit over 10 chassis, each with 4 100Gb uplinks. They don't need to use all of them, but I drool at the thought of a switch or router with 40x100Gb ports. I understand that my ISP is not going to have a 4Tb Internet connection, but the internal network has the potential of a lot of inter-chassis bandwidth.
Here's one way: http://doc.pfsense.org/index.php/Using_IPv6_on_2.1_with_a_Tunnel_Broker
Since pfSense runs on the FreeBSD stable point releases, it tends to be behind the current, which means pfSense was a little slow to the IPv6 party, but is getting lots of support now.
Many DOCSIS2.0 cable modems and all DOCSIS3.0. The top few cable companies, which represent almost 70% of the USA's broadband base, are nearly done with their DOCSIS 3.0 rollout. Rule of thumb, if it supports more than 30Mb, it is IPv6.
Good news, the Internet backbone has been IPv6 for over 10 years now and Cable modem and DSL hardware in the past 5-8 years have all supported IPv6 natively also.
As for "most software". Well, the most commonly used software is the web-browser, which has been IPv6 for a while now also. Most people purchase new hardware on a 5-8 year cycle and nearly all networking hardware in the past few years has been IPv6.
All 4g smartphones are IPv6, it is a requirement for 4g.
10% is the tipping point when a new technology "explodes" is use. So by year 4/5, IPv6 will be nearly everywhere.
NAT does not significantly increase security
^^
Paint on a tank, it makes the tank more resistant to attack. Think of all of the energy dissipated when the paint flakes off from a rocket hitting the armor!
NAT = Paint
Firewall = Armor
This has been impressively ameliorated by the use of NAT, which shares numerous intenral and protected IP addresses behind a single or pair of public addresses
Yet another person who doesn't understand NAT. How do these people get jobs in networking?!
If we consumed 1,000,000 /64s per second, it would take about 585,000 years to deplete IPv6. That is 1.8447x10^25 ip address per second for 580 millennia.
My ISP didn't give me a router, they gave me Gigabit Ethernet Switch+Bridge+Fiber Transceiver. It's like a big LAN, where each customer gets their own vLAN with a gateway which is the chassis. Except, instead of a regular 48 port switch, it's a 480 port Layer3 IPv4/IPv6 chassis with a 2Tb backplane and 4x100Gb uplinks.
My question is what do they plug those uplink ports into. Each chassis can only support 480 people and there are thousands. What does an aggregate switch for 100Gb ports look like? And what router accepts 100Gb ports?
And it may actually take a 100 days if you actually have to bother with testing it in all sort of scenarios, time zones, browsers en daylight savings time etc.
Send UTC and let the browser format and adjust to the local. VERY easy, almost no testing.