Cat5e and Cat6 cables get a 5Gbps speed boost

Gigabit Ethernet is ubiquitous. If you work in corporate IT, chances are you service at least one site wired with Category 5e or Category 6 cabling running 1000BASE-T. However, as networks expand and add new services, Gigabit Ethernet may start to feel a little congested. Adding or replacing cabling is an expensive proposition, but never fear: the IEEE has just ratified proposal 802.3bz-2016, which lays out the new 2.5GBASE-T and 5GBASE-T specifications. The new specs define a method for delivering 2.5Gbps and 5Gbps over existing Cat5e and Cat6 cables.

Source: Mejdal Rasmussen, via Wikipedia

"But wait, TR," gerbils are already crying. "Cat6 cables support 10GbE!" Yes, over short distances, Category 6 cables can run 10GBASE-T. However, the new standard lets those cables support a 5Gbps data rate across 100-meter runs. 802.3bz equally allows for Category 5e cabling to offer 5Gbps performance "on defined use cases and deployment configurations." For other configurations, Cat5e cabling will be limited to 2.5Gbps, which is neverthless a 150% improvement over standard Gigabit Ethernet.

As far back as 2013, networking vendors looked ahead and saw the toll that high-end Wi-Fi gear would take on their networks. The "wave 2" expansion of 802.11ac has a maximum theoretical throughput of almost 7Gbps, and even first-generation 802.11ac can deliver faster-than-GbE speeds with high-end MIMO gear. Enterprise Wi-Fi installations servicing many users could already step over the gigabit boundary with ease.

To cope with this demand, those vendors formed two competing alliances working toward the goal of achieving higher-than-Gigabit speeds over existing wiring. Today's release of an IEEE standard regarding the technology should prevent any further interoperability concerns that might have been raised by the existence of these two separate groups.

Certain high-end gear—like some of Cisco's Catalyst Ethernet switches—already supports the intermediate speeds. However, most 10GBASE-T equipment will not be able to step down to 2.5GBASE-T or 5GBASE-T. Purchasing new networking equipment is both cheaper and easier than re-wiring structures, though.

Comments closed
    • robertsup
    • 3 years ago

    if it will be compatibile with all hardware on market it will be plus but its not then it only will late 10gb move for desktop market

    • phileasfogg
    • 3 years ago

    >>>>> However, most 10GBASE-T equipment will not be able to step down to 2.5GBASE-T or 5GBASE-T.

    If the PHY (transceiver) supports NBASE-T, then it will support all 5 speeds (100/1000/2.5G/5G/10G).
    The new Intel X550-AT cards support NBASE-T.
    I haven’t checked into the Broadcom 10GBase-T PHYs but I expect they support NBASE-T as well.

      • davidbowser
      • 3 years ago

      Intel was in the NBASE-T alliance. Broadcom was in the MGBASE-T alliance. If you read the press releases, the NBASE-T members are claiming software updates will enable the new IEEE standard on NBASE-T compatible chips, but the MGBASE-T alliance is not making the same claim.

      • curtisb
      • 3 years ago

      It’ll be the MAC, not the PHY. 🙂

    • Krogoth
    • 3 years ago

    UTP has no head-room left. You either have to go optical or STP if you want more bandwidth.

    If you have growing bandwidth needs then it makes more sense upgrade everything to STP or optical in the long-run.

    • Bensam123
    • 3 years ago

    Absolutely silly that we’ve been sitting on gigabit for over a decade and a half. I remember buying and I’m still using my first gigabit switch in 99 back when I had to buy a $90 NIC to get gigabit on each computer. Really wish there was a entry there for 10Gbit, but no one ever made it.

    All just seems to be that they’re trying to keep the consumer market and enterprise market artificially separated, just like they still sell 100Mbit switches with guts of gigabit switches and flash different firmware to them.

      • travbrad
      • 3 years ago

      I agree there is definitely some artificial market segmentation going on there but I think there’s also a lack of widespread demand for it on home networking stuff. Consumer/home computing in general has been headed towards lowest common denominator and being “good enough” for awhile. Even gigabit is still faster than a lot of people’s hard drives (and most people don’t transfer anything over the LAN anyway, it’s just their connection to the internet)

      • HERETIC
      • 3 years ago

      It’s not so much the decade and a half on Gb I find absolutely silly but that the industry
      insists on going up in multiplies of ten.

      In my opinion,what would have been wonderful-if several years ago they jumped to 2 Gb
      90% of home consumers would then have found their HD the bottleneck-fastest I have is
      around 200 MB on outside of platters…………………………….

      • Krogoth
      • 3 years ago

      It is because the masses never liked wired ethernet in the first place. It was a foregone conclusion that wireless ethernet would take over in the mainstream when the masses start to moving towards portable systems. It also helps doesn’t help that there’s no killer mainstream application until recently that made 100Mbps woefully inadequate.

      • End User
      • 3 years ago

      [quote<]Absolutely silly that we've been sitting on gigabit for over a decade and a half.[/quote<] Exactly. While wireless tech has improved by leaps and bounds the industry has let ethernet stagnate. Any consumer with a wired home, from roughly 2000 onward (5e), could have benefited from improvements to ethernet bandwidth over that timeframe. I'd kill for a 5Gbps consumer NAS on a 5Gbps consumer network at home.

        • Krogoth
        • 3 years ago

        Wired Ethernet hasn’t stagnated though. It goes up to 40/100Gbps over STP/Optical media. The problem is the masses never cared for wired ethernet in first place and opt for wireless ethernet (despite the performance, reliability and security issues) to go along with their portable systems. It doesn’t help that internet connection speed for the masses only exceeded 100Mbps in recent years and that hasn’t been a mainstream killer app that render 1Gbps Ethernet and all existing wireless solutions woefully inadequate.

        The networking industry simply adjusted to the demands of the market. They structured their wired 10Gbps and beyond solutions towards the professional and enterpirse markets. While shifted towards wireless ethernet solutions for the masses.

          • End User
          • 3 years ago

          masses ≠ all consumers

          There are different levels of consumers. Please try and keep up.

            • Krogoth
            • 3 years ago

            There have been 10Gbps ethernet solutions and beyond for years. The issue you have is that you “feel” that price points for 10Gigabit and beyond is too high to justify the upgrade cost. Not every prosumer feels this way. This new “5Gbps spec” is meant for “you” and “SMB-types” that are in the same position.

            • End User
            • 3 years ago

            [quote<]This new "5Gbps spec" is meant for "you" and "SMB-types" that are in the same position.[/quote<] Thank you for telling us what is so painfully obvious. 5Gbps has been a long time coming. Development was glacial. Thank goodness ARM CPU development is on a faster track.

            • Krogoth
            • 3 years ago

            The problem is that UTP (cheapest networking media) has a number of issues when trying to go beyond 10Gbps. This “5Gbps spec” is the last bump for UTP before we need to over optical/STP.

            It is going to be the PATA-150 of the networking world.

            • End User
            • 3 years ago

            Slow consumer SSDs are 500 MB/s these days. A 100 MB/s network connection is a major bottleneck. I’ll take any bump in network performance I can get at this point.

            • Krogoth
            • 3 years ago

            Actually mainstream SATA SSD media is closer to ~250-300MiB/s sustained depending on the workload and controller in question. They can fall down to a lowly 80-100MiBs if the drive is old and used-up. That 400-500MiB/s figure is only found on higher-end 2.5″ SATA SSD models while M.2 NVMe SSDs easily exceed 1GiB/s.

            • End User
            • 3 years ago

            [quote<]Actually mainstream SATA SSD media is closer to ~250-300MiB/s [/quote<] With cheap consumer SSDs providing double to triple the bandwidth of 1Gbps ethernet even the low-end consumer is dealing with a tremendous bottleneck in network performance. All hail 5Gbps ethernet.

            • Krogoth
            • 3 years ago

            Not really, the mainstream market doesn’t move that much data around on their network on a frequent basis. They usually opt for an external HDD via USB 2/3 for a back-up solution. They are bottleneck by their wireless ethernet solution if anything else.

            It is a different story for prosumers and professionals though.

            • End User
            • 3 years ago

            Excellent. We can exclude the mainstream consumers from any discussion on 5Gbps ethernet or any other high performance consumer hardware discussion for that matter.

            • Bensam123
            • 3 years ago

            If you compare entry today for 10Gbit compared to entry 15 years ago for Gigabit, I could but a $200 switch and a $90 nic… How much does the cheapest 10Gbit setup cost? Triple that? 10G has been around for like a decade too. Still no entry level pricepoints.

            Inflation and all that jazz, etc.

            • Krogoth
            • 3 years ago

            The three forces that prevented 10Gbps Ethernet from coming down in cost and being widespread are

            1.) UTP media has issues trying to transmit data reliably at 10Gbps and beyond with lengths longer than 10 meters. It is by far the cheapest and most widespread networking media. STP costs more and is fragile to work with. Optical increases the problems of both issues by another an order of a magnitude but has far more headroom (theoretically can handle PiBs of bandwidth). 100Mbps and Gigabit Ethernet never had these issues its inception (It worked on existing and affordable CAT5e without an hitch).

            2.) There’s no mainstream demand for anything beyond 1Gbps Ethernet. The markets that have a demand for it are more then willing to pay for the cost and headaches associated with 10Gbps Ethernet and beyond. This new “5Gbps” spec is really meant for SMB demographic where 1Gbps isn’t enough but cost and headaches involved in 10Gbps Ethernet and beyond are too much for their budget. I personally think it is really just a reaction to Thunderbolt 3 and other low-cost ultra-high bandwidth solutions that are on the market.

            3.) Wireless Ethernet has got far more traction in the mainstream market. The masses prefer simplicity and convenience of Wireless Ethernet despite the obvious performance and security issues that it is plagued with. Masses also prefer portable platforms where Wired Ethernet usage is awkward at best.

            • Bensam123
            • 3 years ago

            You’re talking technical data points. If you’re familiar with ethernet you know that it has fallback. If the cable isn’t capable of transmiting data, it goes into a lesser mode all the way down to 10Mbit half duplex in a worst case scenario… on top of a lot of redundancy.

            There is really no reason that this should be a issue, let alone one where people buy the product anyway even if it doesn’t work.

            Entry level pricepoints aren’t about mainstream demand. They never have and never will be. That’s why people are willing to pay a premium for something that isn’t mainstream, just with gigabit when it first arrived at entry level pricepoints. That aside there are plenty of people who want greater then 1Gbit speed. You can simply look at 802.11AC and see that isn’t necessarily a driving point of adoption anyway and that there is a demand there.

            As with all ethernet standards, you plug it in and it plays. This isn’t even a point. There is no added complexity to buying a gigabit switch over a non gigabit switch.

            • curtisb
            • 3 years ago

            [quote<]UTP media has issues trying to transmit data reliably at 10Gbps[/quote<] Not true at all. CAT6 is limited in the distances it can run 10Gbps, but CAT7 can go the full 100 meters. CAT7a can go 40Gbps at 50 meters, and 100Gbps at 15 meters. The caveat with CAT7 is that it's not recognized standard, and is not backwards compatible. CAT7 is also not UTP, but it's also not a traditional STP, either. Traditional STP has a single shield around all pairs. In CAT7, each individual wire is shielded...which adds to the complexity of making, running, and terminating the cables, and is another contributing factor as to why it was never ratified or very widely used. CAT8 will be the next recognized standard, and will allow for 40Gbps at 100 meters, aka 40GBASE-T. You should also be careful talking about optical media as a general term. It's not all created equal. Single-mode fiber will currently go to 100Gbps (it's been tested at MUCH faster, but official specs only go to 100Gbps). Advances in optics allow for 10Gbps full duplex on a single strand (not pair, strand). 40Gbps full duplex can be done on a pair now. This used to require bonding four 10Gbps links, or eight total strands of fiber. There are only two different types of single mode, and it has more to do with signal attenuation over distance. With regards to multimode, though, there are multiple different types. You have 62.5 micron which can only be rated at OM1 and is only reliable for up to 1Gbps. 50 micron can be rated at OM2 through OM4. In order to get 10Gbps it has to be 50 micron rated at either OM3 or OM4.

            • Krogoth
            • 3 years ago

            CAT8 is going to be STP in some shape or form. That’s kinda the point that I was making. UTP is pretty much at the end of line as far as bandwidth headroom is concerned.

            Optical has a ton of bandwidth headroom and its potential hasn’t been fully tapped into yet. I’m aware that optical comes in many forms each with their own caveats. The main reasons behind current use of optical media is distance, resistance to RFI and power consumption.

            • curtisb
            • 3 years ago

            I’m not going to bother debating something that neither of us know for sure. What I do know is that every iteration of copper for Ethernet that has been released has involved new and different ways to twist the pairs, and segregate them within the jacket to prevent cross talk. If they’ve figured out 5Gbps on CAT6 at 100 meters, I can see CAT8 UTP reaching at least 10Gbps at 100 meters. And now that we’re not locked into multiples of 10 anymore, it could be 12.5Gbps, 15Gbps, 17.5Gbps, etc.

            • Krogoth
            • 3 years ago

            You need “shielding” because the frequency on the cabling is getting into the UHF part of the radio spectrum which is plagued by more RFI and cross-talking becomes more of an issue.

            • curtisb
            • 3 years ago

            Yeah, that’s not going to be as much of an issue as you think. You get MUCH more potential crosstalk and interference from the cables themselves being in bundles than from ambient radio waves. You’ll get more from running a cable right across the top of a fluorescent light fixture…

            • Klimax
            • 3 years ago

            But you and others are in tiny portion of market. There is no economy of scale and problem area is fairly large and hard and expensive.

            • End User
            • 3 years ago

            If Krogoth is correct then mainstream suckers have funded the high end tech. Thanks a bunch!

      • Bensam123
      • 3 years ago

      Shit why don’t they do something like MIMO for gigabit? You can do teaming right now but it’s a huge PITA. There should be native support for running multiple ethernet jacks from the same computer to the same switch and have it intelligently just work. Plug and play. How many high end boards have multiple ethernet jacks? How many people have taken time to setup teaming on it? If you have, how often does it just break your internet connection in general (my experience).

      That’s just teaming you can do right now, they could totally make ethernet jacks smaller, compact them, add more terminals, increase strands, issue a new version of cat, and have it be backwards compatible a-la USB.

      They could also do something with compression.

      It’s silly there are a lot of ways of increasing the speed without just increasing ‘raw’ throughput and no one has taken the initiative to do it.

        • robertsup
        • 3 years ago

        link agregation in civil version of windows work only with intel nics, on linux and windows server you can buy for example microtic 4xgígabit nic pcie card but rest of you network must support it, in chepest wariant you must buy 150-200 usd 16-24 port gigabit switch and even with it you ips is only 1gigabit (i know not many have that speed but tv operators fiber modems give only one rj45, sfp+ is a way to expensive for home network

      • ptsant
      • 3 years ago

      I was recently reading about getting a 10GBps-proof installation in my parents house. Since this involves opening holes in concrete, I wanted to make it future proof.

      The thing is, you actually run into some really hard limits on what can pass through copper at certain lengths. Properly installing 10GBps cabling is not trivial: the cables are considerably thicker, you have to avoid right angles, too much bending, you must test the plugs and actually measure the bandwidth (in Mhz) of the installation to make sure you did everything right.

      Fiber is more robust, but as the article explains, people already have Cat 6, so the cost of rewiring is huge.

        • Bensam123
        • 3 years ago

        This is no different then switching from cat5e to cat6. That would also depend heavily on cable length.

        Even then there are plenty of people that run gigabit over cat5. Just because something isn’t in a absolute best case scenario doesn’t mean it wont work.

        Fiber is never going to catch on in the mainstream as long as you have to buy a $2000 fuser to make ends.

    • Chrispy_
    • 3 years ago

    Consumer ethernet has been stuck at 1Gb/s for far too long.

    When will 5GBASE-T be affordable on switches and NICs?

      • willmore
      • 3 years ago

      Soon. I’ve seen SoCs and networking related chips being announced. I’d guess we’ll see stuff by the holidays.

      • Krogoth
      • 3 years ago

      There’s no killer mainstream application that demands more than what 1Gbps can provide. This spec is really meant for SMB-types that need a bandwidth boost but the cost of upgrading to 10Gbps and beyond exceeds the budget.

        • BurntMyBacon
        • 3 years ago

        [quote<]There's no killer mainstream application that demands more than what 1Gbps can provide.[/quote<]Two points of interest here. 1) Just because no "killer" mainstream application demands more bandwidth, doesn't mean that there aren't any that can't make use of it. 2) Some people don't care about the "killer" application. Some people only utilize their bandwidth 0.27% of the time, but when they do use it, they don't want to be kept waiting. There are even plenty of mainstream consumers that exhibit this trait of impatience. That all said, I find that sometimes applications follow the tech. If 10G ethernet were cheap, many people would have it and companies would find a killer application to sell you. I can think of several off the top of my head, but have no desire to argue the merits of undeveloped applications against a theoretical market. [quote<]This spec is really meant for SMB-types that need a bandwidth boost but the cost of upgrading to 10Gbps and beyond exceeds the budget.[/quote<]Sure is. That is the existing market that manufacturers are banking on to bring returns on their investment. This does not, however, preclude professionals, prosumers, and (if the price gets low enough) mainstream consumers from purchasing these products should they decide it meets their needs. You mention wireless being the preferred method of connectivity for mainstream consumers. If we accept this as fact, then we still have a couple of situations where faster than 1Gbps could eventually be useful to them. I submit for your consideration, 802.11ac and beyond. In a world of mainstream users that (by your definition) prefer portable devices and connect mostly to use the internet, there is no killer application that demands 802.11ac [b<]speeds[/b<]. Given the average internet speeds around the world, you're just throwing all that bandwidth down the toilet. Useful upgrades like MU-MIMO could just as easily have been implemented at 802.11n speeds. Yet, 802.11ac does exist and people demand it. As more households are dropping their desktop and even laptop devices, printing and data storage models have changed significantly. Where people used to use external HDDs and direct attach printers, they now need networked printers (usually wireless) and storage (often times cloud). I am now seeing what may be the start of a movement in part back to local (read not cloud) storage. Slow access, high profile security breaches, lack of privacy, and sensitivity of data are all reasons I've been given when a mainstream consumer asks me about setting up a consumer grade NAS. When they point out that their wireless speed is "bigger" than the NAS speed, I explain that the HDDs are slower than both. However, their response is almost always, "Well, how do I make the HDDs faster". I suspect that as soon as prices are practical, some of these mainstream consumers will be interested in purchasing a consumer grade NAS with flash based storage. When this time comes, I hope that they have the good sense to equip it with a link faster than 1Gb ethernet.

          • Krogoth
          • 3 years ago

          You aren’t understanding why Wireless Ethernet has taken over in the mainstream market. The masses never liked Ethernet wiring draping around in the first place. They aren’t interested in drilling holes in walls, adding wallplates and spending the time/effort to tie up a “wired” setup.

          The answer to upcoming bandwidth needs for “mainstream” NAS devices(what an oxymoron) is going to be faster Wireless Ethernet specs. There’s a demand for it so networking guys are going to spend the R&D for it.

          Killer applications are critical for widespread adoption of technology/standards. It also provides price pressures and economic incentive to make technology as affordable as possible so it can saturate the market. If there’s nothing that demands it then the technology/standards is going to stuck in its own niches. We have already seen this with 10Gbps Ethernet and beyond. 10Gbps spec is almost a decade old. There are other countless examples.

      • mFvwv0zduc
      • 3 years ago

      1Gb/s for too long? I wish 😀
      Mate, I still got “newest” router from my ISP and it got 100 Mb/s ports. It’s Huawei router. I know I can spend £50 or even £100 to get better one, but that’s the default, free router from my ISP. I live in UK.

    • End User
    • 3 years ago

    [quote<]Gigabit Ethernet may start to feel a little congested.[/quote<] It has been that way for a decade.

      • Krogoth
      • 3 years ago

      Only for professionals and datacenter environments. They already made the move to 10Gbps and beyond. This new spec isn’t really meant for them.

      The masses prefer wireless ethernet. Just take a look at the networking shelf space at any B&M store and customer-tier etailer. It speaks for itself.

        • End User
        • 3 years ago

        The masses ≠ all consumers (a fact you fail to grasp or deliberately ignore). Any consumer with a QNAP or Synology NAS will benefit from the switch to 5Gbps ethernet.

          • Krogoth
          • 3 years ago

          a.k.a professionals and prosumers.

          You are confusing the needs and demands of prosumers with mainstream customers. This new spec is geared towards SMB-types where gigabit ethernet isn’t fast enough while 10Gbps Ethernet and faster is beyond their budget.

            • End User
            • 3 years ago

            [quote<]where gigabit ethernet isn't fast enough [/quote<] As a consumer I agree with you 100%.

            • Krogoth
            • 3 years ago

            IIRC, you use Titans and Socket-2011 solutions for real-world work with gaming/casual stuff on the side.

            That’s a classical “prosumer”.

            • End User
            • 3 years ago

            GTX 1080
            Ivy Bridge
            Gaming only

            Not a “prosumer” in sight yet I would love it if my home network was based on a 5Gbps backbone.

            • Klimax
            • 3 years ago

            And how much of it is used most of the time? It doesn’t look like you have real need for it, just want.

            On the other hand, this might help my semi-SMB* networks.

            *-small company (10 computers) with some of characteristics of big. Fun…

            • VincentHanna
            • 3 years ago

            You are defining your terms poorly to create an argument here, for no reason. professionals and pro-sumers are not enterprise customers. They purchase and rely upon consumer grade hardware, which in turn makes them a subset of the larger group ‘consumer’. If professionals and prosumers are finding that their needs are not being met, then so too are a portion of consumers saying the same thing. The argument that they aren’t consumers (again, by [b<]definition[/b<], not by the merit of your argument), because they evaluate products with a more discerning/specified/niche usecase in mind is faulty. That, my friend, is a universal condition of the free market. No true scotsman would put a file server in their basement! (removed the car analogy; edited for clarity)

    • xeridea
    • 3 years ago

    [quote<]The "wave 2" expansion of 802.11ac has a maximum theoretical throughput of almost 7Gbps, and even first-generation 802.11ac can deliver faster-than-GbE speeds with high-end MIMO gear.[/quote<] Anyone know the real world max throughput on WiFi routers?

      • End User
      • 3 years ago

      I’ve seen a max of 90 MB/s on my ac network.

        • willmore
        • 3 years ago

        I get 50MB/s, but my network is 2 stream by 80MHz. Going to three more streams and 160MHz will bump that up quite nicely.

        Edited: typo

          • travbrad
          • 3 years ago

          The only problem is bumping up the number of streams and the width of those streams usually makes your latency spike all over the place. Not an issue for web browsing, general use, or large file transfers but gaming is actually pretty acceptable on narrow band 5ghz 802.11ac, while with wider bands it’s just as bad as 2.4ghz (or even worse sometimes)

      • tipoo
      • 3 years ago

      About half of the advertised as usual. Gigabit wifi = 500Mbit in a common house, which in turn is up to 70MB/s.

      • danazar
      • 3 years ago

      According to a study I just conducted in my pre-war apartment building, running SpeedTest on the 3 out of 74 networks in range that were open, about 2Mbps.

      • cygnus1
      • 3 years ago

      My AC gear often achieves over 500mb/s transfers in a noisy RF environment. It’s basically first gen AC gear too with a rated peak of 867mb/s. 5GHz spectrum does a lot to get that real world speed a lot closer to the max theoretical speed as compared to 2.4GHz. I definitely want to upgrade to something that can handle 4 streams and MU-MIMO, once it comes out. All of the 5e runs in my house are well under 55m, so I would certainly upgrade to faster switches once this becomes a real option because this 2.5G and 5G gear is supposed to be a lot cheaper than 10GBASE-T and faster gear.

      • clocks
      • 3 years ago

      Downstairs, I max out at about 180meg. But if I put the laptop 5 feet from the router, I have seen 560meg. FWIW – I just got docsis 3.1 1gb cable.

        • xeridea
        • 3 years ago

        Hmm, so realistically gigabit is fine for wireless, and 2.5Gbit may improve max real throughput in rare cases where everyone has good signal sitting next to router. I only have N, max I get is ~100Mbit, but signal penetration is horrible @ 5Ghz, so 2.4Ghz is typically much faster unless I sit next to router.

          • clocks
          • 3 years ago

          Yeah, I get a stronger connection with 2.4ghz, but those speeds top out under 100meg.

          With Comcasts new XI Docsis 3.1 Gigabit product they provide both a modem, and a Netgear R8000. The router was top of the line a couple years ago, but it lacks newer tech like MU-MIMO.

          At some point I may be tempted to buy something newer with MU-MIMO, and see if it helps me better utilize my 1gb connection.

      • Vaughn
      • 3 years ago

      i’ve yet to see AC pull over 500Mbps.

      however that was with a 2×2 device and not 3×3 but I think this will still be under 1Gbps.

    • meerkt
    • 3 years ago

    Maybe, as a first step, we could finally have router manufacturers stop with their 100Mbps-everywhere nonsense?

      • just brew it!
      • 3 years ago

      Yeah, it’s a little annoying that so many routers still incorporate a 100Mb switch. But if you’ve got more than a trivial amount of wired Ethernet gear you need a separate switch anyway, so it’s not the end of the world.

        • meerkt
        • 3 years ago

        I don’t normally need more than 4 LAN ports.

        I suspect the thing that’s going to drive the disappearance of 100Mb ports will be the increasing speeds of the WAN connection, not the already faster LAN-side speeds. But it might take a few more years.

      • Krogoth
      • 3 years ago

      It is because the majority of UTP cabling out there in the mainstream market still have CAT5 and followed by CAT5e. CAT5e can only do 1Gbps and supposedly this new spec only if the cabling isn’t suspect or isn’t in a noisy RFI environment.

    • TwistedKestrel
    • 3 years ago

    So is this going to work out better than the glacial uptake of 10GbE?

      • End User
      • 3 years ago

      Hopefully we will start to see 5GbE switches soon.

      • danazar
      • 3 years ago

      Well, it has the following advantages over 10GbE:

      1) Hardware that includes it should use less energy and generate less heat than 10GbE products

      2) Integrating it should be cheaper, partly due to that “less energy and heat” thing

      3) Consumers/SMBs will be receptive to “(probably) works with your existing cables”

      4) Wi-Fi now theoretically exceeds GbE speeds, and router makers who add 2.5GbE ports will get a new round of “upgrade your router to achieve higher speeds!”

      So yeah, I think we could see faster adoption here.

        • cygnus1
        • 3 years ago

        [quote<] 4) Wi-Fi now theoretically exceeds GbE speeds, and router makers who add 2.5GbE ports will get a new round of "upgrade your router to achieve higher speeds!" [/quote<] I know this will happen, but it's really dumb. Most people with wifi at home only use the wifi to connect to the internet, not to other devices in the house. And when they do connect to other devices it's usually for screencasting or maybe printing. A whole lot of nothing that will benefit from their router having faster Ethernet ports.

          • yuhong
          • 3 years ago

          I think this is part of why it was created in the first place though. Not for single router environments, but multiple ones.

            • cygnus1
            • 3 years ago

            how would having multiple routers in an environment create a need for faster ethernet links?

            • yuhong
            • 3 years ago

            Because you need a way to connect the multiple routers together.

            • cygnus1
            • 3 years ago

            Wat??

            Do you mean for multiple home internet connections or something? Or are you talking about in enterprise type networks where they have many different VLANs and whatnot and already have 10gb ethernet??

      • Krogoth
      • 3 years ago

      This is meant to be a low-cost alternative to increase bandwidth for some SMB environments. The masses are opting for wireless ethernet solutions over wired ethernet.

    • techguy
    • 3 years ago

    I’m assuming this will require firmware updates in order for existing devices to support the new standard. Confirm/deny?

      • drfish
      • 3 years ago

      I think it’s more along the lines of your cables getting a new lease on life and your networking gear getting replaced.

        • Waco
        • 3 years ago

        This. I have lots of Cat5e in place already and I’d love to be able to continue using it without pulling new cables.

        That is, assuming it gets adopted. 10Gbps has been absurdly slow in dropping in price.

      • RAGEPRO
      • 3 years ago

      Based on the research I did when writing this piece, it seems there are significant differences in how gigabit Ethernet vs. 10 gigabit Ethernet work at a physical level. As a result, I think 10GbE devices could support this with a firmware update, but GbE devices will have to be replaced.

        • willmore
        • 3 years ago

        You’re correct, this requires all new hardware for “anything but the wire”.

        The new speeds are–as the really nice graph shows–basically nerfed versions of 10GigE–slower signaling to fit in with the lower specs of the cables at longer distances.

        I’ve already seen chipsets and SoCs using this start to be announced. I’d expect gear by the holiday season.

        • just brew it!
        • 3 years ago

        I think the modulation scheme is different from 10gbE as well; this may mean that 10gbE gear isn’t compatible either, regardless of firmware upgrades. Not that it matters much; the small percentage of organizations which already have 10gbE gear have obviously already invested the time and money to upgrade their cabling.

Pin It on Pinterest

Share This