Latest Thread Images

Placebophile Cat 8 Ethernet Cable
Dec 15, 2016 at 8:04 AM Post #16 of 54
I asked @watchnerd who were his friends, so if I deleted your post, he hates you.
biggrin.gif
 
seriously I left the first page and removed the flying spaghetti monster that was the rest(sorry, no religion). I was tempted to sink the all topic but watchnerd seems to care a little. so here we are.
 
please try to stick to this ethernet cable, or at least to ethernet cables in general or stuff that actually have the same use. or create your own topic to talk about whatever it is you want to say. it's not that hard.
thank you, happy new year.
 
Dec 27, 2016 at 10:08 AM Post #17 of 54
chaos215bar2 said:
 
The trouble with wireless is, there's a fixed amount of bandwidth available to be shared by all devices and services operating in a given area, and the only way to get around this is to place access points (be that Wi-Fi, LTE, or something else entirely) closer together, reducing their power output so they don't interfere. Even now, completely saturating the available 2.4 and 5 GHz bands, Wi-Fi struggles to reach 1Gb/s of realistic throughput — something trivially achieved by wired technology that's ubiquitous and nearly two decades old.
 
Yes, I think that any ISP with the infrastructure in place would love to see internet access go fully wireless, but not for the reasons you propose. Wireless offers a great opportunity to bypass some of the regulations (in the US, anyway) surrounding wired internet access and to charge hefty overage fees on meager data caps — unless of course you use the provider's own content store, which will naturally be exempt from caps.
Click to expand...

 
I think the thing being missed in this statement is that the wireless spectrum is limited in the same sense the optical spectrum is.  There is a finite range but the use of it depends on how well you can divide it up and how narrow a guard band you have to maintain on either side of the active channel to prevent crosstalk and bleed-over.   What is changing rapidly is the width of the channels and that process will free up more and more bandwidth as older devices age out and newer more efficient devices replace them.   A few years ago if you and your neighbor had cordless phones there was a good chance you could hear his calls superimposed on yours.  Now, that has improved considerably and I would expect we are still at least a couple orders of magnitude away from the absolute limits of what can be done.    This is in a lot of ways similar to what we have seen in optical technology.  Early laser optics were one color, then they figured out that you could transmit in several distinct wavelengths over the same cable and improve bandwidth, now we increasingly see those frequency divisions getting narrower and more channels being available for data being sent over the same cable.    In many ways, optical and wireless have more room to grow than does wired technology.  We know the theoretical limits of electrical signal conduction and moving multiple data streams over copper wire is achieved either by the addition of additional pairs of wires or by some form of multiplexing which simply divides what is available between streams but does not free up any additional usable capacity.
 
Dec 31, 2016 at 11:08 AM Post #18 of 54
wiljen said:
   
I think the thing being missed in this statement is that the wireless spectrum is limited in the same sense the optical spectrum is.  There is a finite range but the use of it depends on how well you can divide it up and how narrow a guard band you have to maintain on either side of the active channel to prevent crosstalk and bleed-over.   What is changing rapidly is the width of the channels and that process will free up more and more bandwidth as older devices age out and newer more efficient devices replace them.   A few years ago if you and your neighbor had cordless phones there was a good chance you could hear his calls superimposed on yours.  Now, that has improved considerably and I would expect we are still at least a couple orders of magnitude away from the absolute limits of what can be done.    This is in a lot of ways similar to what we have seen in optical technology.  Early laser optics were one color, then they figured out that you could transmit in several distinct wavelengths over the same cable and improve bandwidth, now we increasingly see those frequency divisions getting narrower and more channels being available for data being sent over the same cable.    In many ways, optical and wireless have more room to grow than does wired technology.  We know the theoretical limits of electrical signal conduction and moving multiple data streams over copper wire is achieved either by the addition of additional pairs of wires or by some form of multiplexing which simply divides what is available between streams but does not free up any additional usable capacity.
Click to expand...


While it is true that older, bandwidth-inefficient protocols are being phased out, making room in the RF spectrum for newer, more efficient ones, freeing bandwidth is not as simple as reducing the width of the communications channel.
 
To start, the fact that cordless phones (or any other wireless communication protocol these days) don't generally interfere in any obvious manner, such as superimposing two conversations, has little to do with using more narrow communications channels. One byproduct of digital communications is that two devices can, fairly easily, make use of the same frequency spectrum in a time-multiplexed manner (taking turns transmitting) and reject and retransmit anything lost due to interference from overlapping transmissions or outside sources. This is actually critically important to high data rate wireless protocols such as 802.11ac, because these protocols rely in increasingly wide channels to achieve their maximum rates. Yes, you can achieve ~2 GB/s with 802.11ac, but there are only two non-overlapping 160 MHz channels in the 5 GHz band to support that.
 
More generally, data rate is theoretically bounded by channel bandwidth and transmit power. (See here and here for some sources that might start to provide a sense of why this is.) There's certainly still room for improvement, but you can't simply divide the RF spectrum into ever smaller channels to achieve unlimited communications bandwidth. Ultimately, RF, copper (which is essentially just RF with a waveguide), and optical communications all follow the same rules. The differences come down to the physics that affect transmission (e.g. signal bandwidth and noise) in each range of the electromagnetic spectrum.
 
Dec 31, 2016 at 11:20 AM Post #19 of 54
chaos215bar2 said:
 
While it is true that older, bandwidth-inefficient protocols are being phased out, making room in the RF spectrum for newer, more efficient ones, freeing bandwidth is not as simple as reducing the width of the communications channel.
 
To start, the fact that cordless phones (or any other wireless communication protocol these days) don't generally interfere in any obvious manner, such as superimposing two conversations, has little to do with using more narrow communications channels. One byproduct of digital communications is that two devices can, fairly easily, make use of the same frequency spectrum in a time-multiplexed manner (taking turns transmitting) and reject and retransmit anything lost due to interference from overlapping transmissions or outside sources. This is actually critically important to high data rate wireless protocols such as 802.11ac, because these protocols rely in increasingly wide channels to achieve their maximum rates. Yes, you can achieve ~2 GB/s with 802.11ac, but there are only two non-overlapping 160 MHz channels in the 5 GHz band to support that.
 
More generally, data rate is theoretically bounded by channel bandwidth and transmit power. (See here and here for some sources that might start to provide a sense of why this is.) There's certainly still room for improvement, but you can't simply divide the RF spectrum into ever smaller channels to achieve unlimited communications bandwidth. Ultimately, RF, copper (which is essentially just RF with a waveguide), and optical communications all follow the same rules. The differences come down to the physics that affect transmission (e.g. signal bandwidth and noise) in each range of the electromagnetic spectrum.
Click to expand...

 

 
Dec 31, 2016 at 7:43 PM Post #20 of 54
chaos215bar2 said:
 
While it is true that older, bandwidth-inefficient protocols are being phased out, making room in the RF spectrum for newer, more efficient ones, freeing bandwidth is not as simple as reducing the width of the communications channel.
 
To start, the fact that cordless phones (or any other wireless communication protocol these days) don't generally interfere in any obvious manner, such as superimposing two conversations, has little to do with using more narrow communications channels. One byproduct of digital communications is that two devices can, fairly easily, make use of the same frequency spectrum in a time-multiplexed manner (taking turns transmitting) and reject and retransmit anything lost due to interference from overlapping transmissions or outside sources. This is actually critically important to high data rate wireless protocols such as 802.11ac, because these protocols rely in increasingly wide channels to achieve their maximum rates. Yes, you can achieve ~2 GB/s with 802.11ac, but there are only two non-overlapping 160 MHz channels in the 5 GHz band to support that.
 
More generally, data rate is theoretically bounded by channel bandwidth and transmit power. (See here and here for some sources that might start to provide a sense of why this is.) There's certainly still room for improvement, but you can't simply divide the RF spectrum into ever smaller channels to achieve unlimited communications bandwidth. Ultimately, RF, copper (which is essentially just RF with a waveguide), and optical communications all follow the same rules. The differences come down to the physics that affect transmission (e.g. signal bandwidth and noise) in each range of the electromagnetic spectrum.
Click to expand...

Spot-on.  Thank you!
 
Jan 12, 2017 at 2:03 PM Post #21 of 54
Good quality Cat7 should be enough for any power-user's household. I'd be more interested in newly emerging 2.5GBASE and 5GBASE standards than getting uber-expensive 10Gig cards, managed switches, routers,..
 
Jan 12, 2017 at 4:46 PM Post #22 of 54
asymcon said:
  Good quality Cat7 should be enough for any power-user's household. I'd be more interested in newly emerging 2.5GBASE and 5GBASE standards than getting uber-expensive 10Gig cards, managed switches, routers,..
Click to expand...

Total overkill.  Cat6, and in most cases Cat5e is more than adequate for everything in any home and most businesses.  Don't be confused by the so-called bandwidth specs, it's a length issue too, and the typical 50' run can pass much more speed than you'd think just by reading the numbers. Unless somebody's throwing uncompressed 4K around (and nobody is), it's not even a stress.  Far more important than the wire is a switch with a true high-speed backplane.  But who want's to pay for that?  Oh well.
 
Jan 13, 2017 at 12:40 PM Post #23 of 54
pinnahertz said:
  Total overkill.  Cat6, and in most cases Cat5e is more than adequate for everything in any home and most businesses.  Don't be confused by the so-called bandwidth specs, it's a length issue too, and the typical 50' run can pass much more speed than you'd think just by reading the numbers. Unless somebody's throwing uncompressed 4K around (and nobody is), it's not even a stress.  Far more important than the wire is a switch with a true high-speed backplane.  But who want's to pay for that?  Oh well.
Click to expand...

I'm throwing 500MB/s low latency read into the mix and I need that full-duplex without compromise. Upgrade to 10GBASE is not entirely out of question, but not with current HW prices. The problem with 2.5/5GBASE might be with "legacy" interfaces - e.g. NICs might not come as ExpressCard expansions. That's why I'm stuck with 1gbps for now. :frowning2:
 
Jan 13, 2017 at 11:59 PM Post #24 of 54
pinnahertz said:
  Total overkill.  Cat6, and in most cases Cat5e is more than adequate for everything in any home and most businesses.  Don't be confused by the so-called bandwidth specs, it's a length issue too, and the typical 50' run can pass much more speed than you'd think just by reading the numbers. Unless somebody's throwing uncompressed 4K around (and nobody is), it's not even a stress.  Far more important than the wire is a switch with a true high-speed backplane.  But who want's to pay for that?  Oh well.
Click to expand...


If you're installing wiring in a wall, though, why are you looking only at current standards? The price of a good quality Cat6a cable will easily be dwarfed by installation costs, and gives you some headroom if another standard comes along in 10 or 20 years that actually does offer an advantage to the higher specced cable over shorter runs. Even Cat7a isn't that expensive. In the worst case, you'll have increased building costs by a fraction of a percent or installation costs by maybe a few percent, assuming your building isn't already constructed to make data cabling easily accessible and upgradable (which is an entirely different scenario).
 
When I see an ad for a new development touting Cat5e wiring (occasionally curiosity gets the better of me when they arrive in the mail), my first thought is, what else did they cheap out on so they could save a few dollars in building costs on a unit that's going to sell for several hundred thousand dollars at least. If they said Cat6a or Cat7a instead, I would think "Wow, they actually cared about designing a building that's going to last for more than a decade before things start to break down or become hopelessly outdated."
 
Certainly you'll never be able to predict everything that's coming, but why not run a few Cat7a, fiber, and coax drops to each room if you really care about such things. You don't have to hook them all up right away, and that will certainly cover things for the foreseeable future.
 
Jan 14, 2017 at 12:06 AM Post #25 of 54
If you want to future proof a new installation, worrying about the best wire to use now is still not going to do it. But if you pay up and put in wire duct so Cat5/6/7 can be yanked out and used to pull in whatever the flavor of the day is, now you've invested in the future for real.
 
Jan 14, 2017 at 1:01 AM Post #26 of 54
The thing is - why settle with 5/6/6a for household installation, when Cat7 isn't overly expensive and is a future-proof solution? Last time I checked it was $1.5/metre. Even some mains cables are more expensive than that. 
 
Jan 14, 2017 at 1:31 PM Post #27 of 54
asymcon said:
The thing is - why settle with 5/6/6a for household installation, when Cat7 isn't overly expensive and is a future-proof solution? Last time I checked it was $1.5/metre. Even some mains cables are more expensive than that. 
Click to expand...

The installation cost is higher for Cat7/8 for reasons beyond the wire itself. Termination is slower, increasing labor cost. Connectors are much more expensive and harder to find. Certification tools are more expensive. It all adds up, but when the application doesn't even challenge Cat5e or 6, it's hard to justify even to future proof. For homes it's highly unlikely there will be a need for capacity beyond Cat6, as the speed bottleneck is still wifi, and even more so, servers out on the Internet. Already the key performance gains have been shown in advanced codecs that permit higher quality video with minimal bandwidth increase. That trend will likely continue as the difficulty and disadvantages of higher bandwidth media will become more of a factor. Already most ISPs have data caps, and those aren't going away. It all means that bandwidth for media will likely not leap up quickly, and we aren't challenging cat5e even now for single drops of average length. We get 1gig through the full 100m length of Cat5e, more than 10X the speed of the typical fast internet connection. Shorter runs go even faster. Even as we move to fully IP based media distribution, the infrastructure will still hold internet speeds down below LAN speeds for quite some time.

For the future paranoid, I still suggest wire duct, making any possible upgrade easy.
 
Jan 14, 2017 at 1:59 PM Post #28 of 54

pinnahertz said:
The installation cost is higher for Cat7/8 for reasons beyond the wire itself. Termination is slower, increasing labor cost. Connectors are much more expensive and harder to find. 
Click to expand...

As far as I know, you can use standard FTP connectors meant for shielded 5e. With DIY no cost is increased. Plugs are RJ45 compatible. 
 
 
pinnahertz said:
It all adds up, but when the application doesn't even challenge Cat5e or 6, it's hard to justify even to future proof. For homes it's highly unlikely there will be a need for capacity beyond Cat6, as the speed bottleneck is still wifi, and even more so, servers out on the Internet. 
Click to expand...

Maybe not for Jane and Joe, but me personally, I'd use up the BW very quickly. In fact, I'm currently bottlenecked with 1Gbps for my "server farm" use and as said previously, could use 500MB/s (5Gbps) duplex. 
 
 
pinnahertz said:
more than 10X the speed of the typical fast internet connection
Click to expand...

But there's always LAN communication, and 1Gbps in current age of conventional 550MB/s SSDs, 3.2GB/s M.2 SSDs and ramdisks could be limiting factor, especially for low-latency delivery (<2ms).
 
Jan 14, 2017 at 3:22 PM Post #29 of 54
asymcon said:
pinnahertz said:
The installation cost is higher for Cat7/8 for reasons beyond the wire itself. Termination is slower, increasing labor cost. Connectors are much more expensive and harder to find. 
Click to expand...

As far as I know, you can use standard FTP connectors meant for shielded 5e. With DIY no cost is increased. Plugs are RJ45 compatible. 
 
 
pinnahertz said:
It all adds up, but when the application doesn't even challenge Cat5e or 6, it's hard to justify even to future proof. For homes it's highly unlikely there will be a need for capacity beyond Cat6, as the speed bottleneck is still wifi, and even more so, servers out on the Internet. 
Click to expand...

Maybe not for Jane and Joe, but me personally, I'd use up the BW very quickly. In fact, I'm currently bottlenecked with 1Gbps for my "server farm" use and as said previously, could use 500MB/s (5Gbps) duplex. 
 
 
pinnahertz said:
more than 10X the speed of the typical fast internet connection
Click to expand...

But there's always LAN communication, and 1Gbps in current age of conventional 550MB/s SSDs, 3.2GB/s M.2 SSDs and ramdisks could be limiting factor, especially for low-latency delivery (<2ms).
Click to expand...

Not if you consider the actual media bit rate. No current media requires anything near that speed.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top