atcomsystems.ca/forum Forums VOIP Phones & IP PBX Networking Gigabit to Desktop: Is it a hype or real demand?

Read an article recently: Gigabit to desktop? Not so fast . Although this was an old article back in 2002, I see some points reflected in the article are still valid. I personally think "Gigabit to desktop" is more of a hype from equipment vendors than a real demand from majority of users.

How do think? I know there many people on this forum are professionals and gurus with first-hand customer experience in voice and data network cabling and setups, would love to see their comments on this question.

Have customers that started at 10 then 10/100. They're running gig now and all smiles. There must be a balance of hardware and software to achieve the full benefit.

Gigabit ethernet was an emerging technology in 1997. At this time, a lot of networks were still running 10Base-T. The cost of 100Base-T hardware was coming down at this time, so people were making the switch to 100Base-T just as gigabit gigabit was coming onto the market.

Now, we have 10Gigabit over copper on the market. Gigabit network cards are dirt cheap, and most new computers, certianly servers, come standard with gigabit NICs. Gigabit switches are quite affordable now as well.

You can argue that gig to the desktop is overkill, but that argument is very dated. Even if someone doesn't feel they NEED gigabit to the desktop now, nobody in the right mind would choose to install a technology that would prevent them from upgrading to gigabit later on.

I know you're bringing this up because of your product, but in my honest opinion you are about 10 years too late to market.

10 years ago the the average user wasnt working with 1 gig or larger files. Nowadays its not uncommon to transfer 4 gigs to or from a server and gigabit certainly makes those transfers a lot less painful. Feels kinda silly to respond to a 10 year old speculative argument which really doesnt hold water today. Gigabit to the desktop is pretty common today with most new pcs equipped with gigabit cards and gigabit switches very reasonable priced. I even have gigabit at home. If you had posted an article about 10gig to the desktop then more people may have responded but 1Gig in my opinion is a no brainer

Looks like most of us have been sold the idea of Gigabit to Desktop. I agree gigabit ports becomes very cheap over the years, but it is still at least as twice expensive as 10/100 today.

Sometimes Gigabit to desktop may become more expensive than we original thought. For many networks that are already deployed with 10/100 IP phones, running Gigabit to Ethernet may be a costly decision as this requires replacing 10/100 IP phones with gigabit IP phones, which probably is $100 more expensive per unit. This is equally true for a new network.

Another observation is that most 16 or 24-port gigabit switches available today can only use one of 1G ports as uplink when multiple switches are interlinked becasue a 10G port is still way too expensive to be available. Technically this would cause traffic blocking. Cisco (read it here ) explains that 1G uplink is OK under most situation for interlinking computers and servers because of the burst nature of data traffic. However, I would ask why they build their 10/100 switches with 1G uplink ports, not 10/100 ports? Under certain situations such as multiple PC backup by a server, the 1G uplink of gigabit switches will definitely be the bottleneck.

Therefore, I should say although Gigabit Ethernet to desktop can provide network performance over fast Ethernet at a premium, majority of users actually don't need or ask for such performances in accomplishing their daily work.

Sometimes I would like to take 10 minutes instead 10 second to download a 5GB file, so that I can take a break and get a cup of coffee or something else, which is definitely a much more healthy way to live. :-))

Most customers want Gigabit and when the price of 10 Gig comes down they will want that. Now if they had stuck with a TDM phone system there would be no need to replace it when upgrading the network.

John,
That Cisco whitepaper pretty much made every argument for gigabit to the desktop. It found that typical users can and do benefit from gigabit. You dont have to replace 10/100 phones because you implement a gigabit switch. All my switches are gigabit but my ip phones are 10/100. Most gigabit switches will autonegotiate 10/100/1000. Im not sure I follow your bottleneck scenario. The uplink between switches is a backbone connection and as such its typically a higher speed than the station ports. You can also get them with 10G uplink ports. You want the backbone connection to be as fast as possible so it wouldnt make sense to make them 10/100. If you cant afford a full gigabit switch then you can buy a 10/100 switch with gigabit uplink ports. That way you can at least have a gigabit connection to another switch or to a server or group of servers which are probably going to be running gigabit. What is your affiliation with Dual-Comm?

Reminds me that "your network is only as fast as the slowest bottleneck" routine with customers.

I have a customer with a two locations, 1500 feet apart, with a big building in between the offices. We use a dry pair from the telco and a VDSL modem. It's 25 mbps. Switches in the buildings are 10/100. there's no need to go gig because the bottleneck between the offices is 25 meg.

John, you clearly aren't here to be convinced of anything. You have a product to sell, and having that link in your signature is going to improve your Google search ranking. Most people come to this forum with honest questions that require expert answers. You came here with a loaded question intended to raise interest in your product. Your arguments against gigabit ethernet are just rationalizations because your "CableShare" switches are limited to 10/100Base-T. I will address some of these points simply to correct your misinformation in this forum.

Does it cost a bit more to purchase gigabit hardware? Yes, but the difference is much smaller than it was even two years ago. You may enjoy the extra coffee break every time you transfer data over the network, but to your boss that is lost productivity. Gigabit may be slightly more expensive, but it can have a much higher ROI.

Your scenario describing a bottleneck with a gigabit uplink makes an interesting point. With 24 or 48 ports running at 100Base-T, that 1000Base-T uplink port can still be a bottleneck during peak network traffic. This is absolutely true, especially in poorly designed and/or poorly managed networks. Your "CableShare" switch on the other hand does not have any port faster than 100Base-T. So that bottleneck in this scenario is a full 10 times worse. On that Cisco switch, and many other managed switches, you can use link aggregation to double your uplink speeds and reduce the bottleneck. Whoops...now the bottleneck is 20 times worse with your switch. You can probably extend your extra coffee break to an entire afternoon off.

In the world of ethernet a splitter of any kind is at best a last resort. I'm sure your switches are a fine alternative when doing things properly just isn't an option...and you don't mind sticking with 100Base-T...and you don't want the option of using POE...and you hate standards and best practices.

Topher, that bottleneck is in between two buildings so that doesnt mean the entire network is going to be experiencing a bottleneck. The intra building communication could still run gigabit and benefit from it. Obviously any inter building traffic would be 25Mbps.