My full colour holographic 1:1 Earth emulator needs this to avoid pixellation.
The 25 Gigabit Ethernet Consortium has decided to ditch its name as it heads off in search of serious speed. The Consortium on Monday announced the change and explained that it was founded to develop 25, 50 and 100Gbps Ethernet, but as it has knocked those off and now has an 800Gpbs spec to share the name doesn’t fit well any …
While I agree that telcos use multiplexing, I disagree with the simple application of that solution to improve throughput.
When the GB standard reached my motherboard NIC, I didn't need to change my Ethernet cable. The same wires work on 100Mbps just as well as on 1Gbps networks.
Fiber optics can have a signal multiplexed to multiply the bandwidth available, but it's still happening in the same glass tube.
So I think that the consortium is working on specs for signals that will allow for 800Gbps in the same wire (probably fiber). Those wires will then be multiplexed to provide for Tbps-capable boxes where you plug in 16 wires or something and boom, you have the enough bandwidth to stream all of the porn at the same time.
Because that's the endgame, right ? Right, I've got my coat.
Ha, the ZX Spectrum did 1500 bps!
I remember 4 Mbps token ring and 10 Mbits "cheapernet".
The coax based 10 Mbps could easily drop to 500 kbps. I demonstrated this to a customer and he upgraded to Cat5. Though a lot of the PCs were still 10 Mbps, it used a switch, not a hub. The 10 Mbps Cat5 with a hub was just as slow as coax, but you didn't have to check the back of every PC to see where the BNC cable had come off the T-piece. We did add plastic T-shells to some offices too mean to upgrade to Cat5. Twenty five years later I'm still using some of the scrapped 50 Ohm cable to make patch cables occasionally for radio gear.
I dumped the last box of swapped out Token Ring ISA cards at the recycling centre only a couple of years ago, with a couple of token ring "hubs" and the giant hermaphroditic auto closing plug//socket patch cables.
Only TV, FM/DAB and Satellite. TV and Sat LNB 75 Ohm cable wants to be rather better stuff than the 75 Ohm equivalent to "cheapernet" which is 50 Ohm like RG58. The 75 Ohm equivalent is RG59, only used for FM radio/DAB now. RG6 / PF100 etc is the usual 75 ohm cable now.
BNC is usually 50 Ohms, though 75 Ohm exist. PL259 is 50 Ohms (CB, Marine radio, Ham Radio). N-type is usually 50 Ohms, but a 75 Ohm version for cable similar size to RG213 does exist.
Almost all other radio applications use 50 Ohms. Though I'd only use RG58 for patch cords or a shortwave receive only cable. The RG213 (much fatter than RG6) 50 Ohm is better for VHF/UHF receivers or transmitters.
Coax ethernet (Vampire Tapped and certainly Cheaper net) was probably pre-existing RF cable?
I've some 92 Ohm cable that was used for some sort of terminal.
It would be most interesting to have estimates of actual performance for real world data transfers. That's frequently quite a bit less than notional line speed. It would also be interesting to told the proposed application. Even if a NIC can keep up, the computer (and application) behind it would have to be very fast indeed.
I doubt the group expects even next-gen bleeding-edge processors to suck on a full pipe. This seems more like a way to get data from a particle physics experiment through a forest of successive aggregating routers to the humongous array of godawful expensive storage located a "safe" distance away. Those gigabytes per nanosecond can add up, ya know.
Hmmm... How much impact on climate change will this link have when it gets applied to transcontinental ocean cables?
Even the fastest network optimised CPUs struggle to move 100Gbps.
This is about uplink aggregation either via network equipment or within blade server (or similar) chassis.
There are big savings for large telcos/cloud providers in being able to uplink a rack via 4x redundant uplinks versus 2-4x as many cables and corresponding switch ports.
These 800G links aren't intended for PC to PC or even LAN links, they're intended for use within and between switches in data centres and in the Internet backbone where huge numbers of lower-speed Ethernet channels (400G, 200G, 100G, 50G, 25G, 10G...) are muxed together. In the longer-haul links (100s, 1000s or 10000s of km) many of these channels are multiplexed together on one optical fiber using different laser frequencies -- just like radio channels, except one channel is typically 50GHz wide (or more), the total "radio bandwidth" is about 5THz centred on 190THz, one fiber can carry getting on for 25Tb/s, and one cable (for example under the Atlantic) might have 8 fibers and be able to carry 200Tb/s.
Before anyone says "this is all pie in the sky, why would anyone want this?" these transceivers are aimed at hitting the market in 2022 when the next generation of switch chips (25.6Tbps in one chip) hit the market, all driven by the apparently unstoppable demand for more bandwidth at lower cost, which in turn is largely down to video streaming -- which is now mainly mainstream film and TV, not porn for once. How do I know all this? Because I've been working on 800G transceiver designs since last year...
But....desktops have gone from 10Mb to 100Mb to 1GB...and stopped.
When are desktops getting 10GB?
1GB/s ~= 100Mb/s, my NVMe SSD is supposed to be faster than that.
If I want to backup my stuff software/pics (currently 570GB) to another desktop it still takes hours!
My NUC has Thunderbolt 3
I tried Thunderbolt 3 (specific card for my motherboard and revision) with a £50 cable (!) and
never got it to work.
Waiting for 10GB to the desktop!
1Gbps is close to what you will get off spinning rust on a desktop.
10Gbps will likely be constrained by your CPU on sun 6-core CPUs and 2.5Gbps/5Gbps "multigig" Ethernet are more likely to match your sustained system throughput unless you use a high-spec desktop fro. The last 2-3 years. Multigig also has the advantage of working with lower spec cables and at much lower power. They just need the switches and NICs to come down in price.
Biting the hand that feeds IT © 1998–2022