So DSL through Verizon is finally available in my area.  I have been using cable modem through Mediacom for the last three years.  Mediacom charges me $56 a month for high speed cable.  Verizon will charge me $44.95 a month.  When I check under Internet Options on my computer, it says the cable speed is 100 MBS...  There's no way it is that fast.  How can I test it?  Here's the rub:

Cable - $56/mo, seems to go out once a month, slows waaaay down during peak hours, 100 MBS????.
DSL - $45/mo, 15 MBS download/2MBS upload

Help a brutha out!

cable modems operate on a shared subscriber "loop" - that is you are sharing bandwidth with potentially a couple 100 or 1000 other subscribers. During peak time you can run into route delays due to bandwidth saturation.

ADSL is considered a point-to-point connection - ie you have a dedicated link back to the ISP but the technology is limited in the total bandwidth available for use.

Other than what you already mentioned its pretty much a draw.  Cable modems tend to be faster on average than ADSL lines but are more sensitive to signal loss on the physical cable segment.  

Cable modems typically use MAC address authentication, ADSL uses PPPoE authentication (user name and password).

Cable speed is not 100mbs.  Your ethernet connection to the modem might be but Cable Modem's upstream typically runs about 3mbs *ideal*, not actual.

Your DSL numbers seem off.  ADSL typically runs 64/128kb up and 256/512kb? down - unless we're talking mega-bits vs. mega-bytes ...

hope this helps.

www.2wire.com has a speed test. Speed meter upper right.  Cable around here is 3megs(3072kbs)  a second. dsl is 320kbs .
cable is $49 mth
dsl is     $30 mth

current test  2973.8

The 100meg/sec your Internet conection sows is actually  the speed between your computer and cable modem. not the actual internet connection. if you want to see how fast you are conecting to the internet, then try here bandwidthplace.com/speedtest/ or other places that actually test how fast your modem to internet conection is.

The general concensous seems to be that cable is ?normally" faster than same priced DSL if your cable company is provisioning the bandwidth correctly and/or not throttling your speeds. Basically if they are doing their jobs, only so many homes have access to your node of the network so that it doesn't get overloaded. If they are not, or say all your neighbors like to download gigs of porn all the time, then your connection may slow down quite a bit. Some cable companies also choke off bandwidth so that you can only download or upload at certain lower speeds which are often slower than comparable DSL rates.

DSL is your own private connection to the Internet, but generally does not have the peak bandwidth as cable for the same price. I personally have a DSL line through Qwest at 1.5 meg down and 786k up for 29.99/month, but I also have to pay 19.99/month for my ISP. ot as fast as cable in my area, but fast enough for me, and I HATE my cable provider.

Quoted: So DSL through Verizon is finally available in my area.  I have been using cable modem through Mediacom for the last three years.  Mediacom charges me $56 a month for high speed cable.  Verizon will charge me $44.95 a month.  When I check under Internet Options on my computer, it says the cable speed is 100 MBS...  There's no way it is that fast.  How can I test it?  Here's the rub: Cable - $56/mo, seems to go out once a month, slows waaaay down during peak hours, 100 MBS????. DSL - $45/mo, 15 MBS download/2MBS upload Help a brutha out!

Those numbers don’t add up.

DSL is usually:

256kbs down and 128 kbs up
Or
1.5 mbs down and 256 kbs up
Or
3 mbs down and 256 (to 512) kbs up

Charter Cable here was either:

384 kbs down and 128 kps up
Or
3 mbs down and 256 kbs up

I just dropped cable for DSL because of the daily slow downs in speed between 3:00pm and 11:00pm due to people getting home, the usual monthly loss of service for 12-36 hours, and DSL was cheaper.

One point on "shared networks" and cable vs. DSL.  The idea that your cable slows down due to heavy usage is silly.  Does the ethernet at your office "slow down" when everyone is there?  Of course not, the bandwidth is pretty much fixed.  The Internet may appear slower due to traffic congestion or server use, but that is not usually your individual connection.  That network in turn is connnected to the IP backbone of the cable company.  Your speed is dependant on how big their transit port is to the big, bad, porno-filled Internet, in other words all the users of the cable company.

All the users of DSL in your area all run to the same DSLAM.  Basically a multi-plexer for your DSL.  So if that particular CO (central office) where the DSLAM is has a 45 meg connection to the actual IP network of Verizon (and by extension the actual Internet), then the congestion model works just as well as the freeway does at rush hour.  As soon as Verizon sells a bunch of DSL connections in your area, BOHICA.

G23c

I just went by price.  Speed by cable is faster though in actual usage I don't notice any real difference.  Where I noticed the difference was in the pocket book.  Because I have cable for the T.V. they gave me a good deal on internet connection.  $20.00/month cheaper than I was paying for Earthlink/Covad DSL.

For most people, they are both fast and it don't matter, go with the best deal.

Also, there is no black and white with this. In one city cable may be much better than dsl, in other cities it could be the opposite.

I've had both. Cable was fast but the service was always dropping, so went to dsl and never had a problem so I'm sticking with that for now.

Quoted: One point on "shared networks" and cable vs. DSL.  The idea that your cable slows down due to heavy usage is silly.  Does the ethernet at your office "slow down" when everyone is there?  Of course not, the bandwidth is pretty much fixed.  The Internet may appear slower due to traffic congestion or server use, but that is not usually your individual connection.  That network in turn is connnected to the IP backbone of the cable company.  Your speed is dependant on how big their transit port is to the big, bad, porno-filled Internet, in other words all the users of the cable company. All the users of DSL in your area all run to the same DSLAM.  Basically a multi-plexer for your DSL.  So if that particular CO (central office) where the DSLAM is has a 45 meg connection to the actual IP network of Verizon (and by extension the actual Internet), then the congestion model works just as well as the freeway does at rush hour.  As soon as Verizon sells a bunch of DSL connections in your area, BOHICA. G23c

Sorry no sell

I have been on lots of shared networks that slowed down as traffic increased. And yes a Ethernet network will absolutely slow down when everyone is there and you exceed throughput capabilities of the network, seen it happen many times. You ever heard of sub-netting, one of the primary reasons to sub-net a local network is to increases efficiency and cut down on slow downs due to increased traffic.

There is absolutely NO DOUBT my cable connection speed dropped EVERY weekday afternoon between 3:00 and 11:00 PM. The reason is obivious. If your cable company puts to many people on one loop YOU WILL SEE SLOW DOWNS.

A competent cable company won't have slowdown.

Quoted: So DSL through Verizon is finally available in my area.  I have been using cable modem through Mediacom for the last three years.  Mediacom charges me $56 a month for high speed cable.  Verizon will charge me $44.95 a month.  When I check under Internet Options on my computer, it says the cable speed is 100 MBS...  There's no way it is that fast.  How can I test it?  Here's the rub: Cable - $56/mo, seems to go out once a month, slows waaaay down during peak hours, 100 MBS????. DSL - $45/mo, 15 MBS download/2MBS upload Help a brutha out!

Your connection TO your cablemodem is probably a 100 mb/s ethernet connection.  That is what Windows is reporting.

Your cablemodem connection is almost guranteed not to be 100 mb/s--more like 10 mb/s, max.  Probably more like 3 to 5 mb/s.

Jim

Quoted: Quoted: One point on "shared networks" and cable vs. DSL.  The idea that your cable slows down due to heavy usage is silly.  Does the ethernet at your office "slow down" when everyone is there?  Of course not, the bandwidth is pretty much fixed.  The Internet may appear slower due to traffic congestion or server use, but that is not usually your individual connection.  That network in turn is connnected to the IP backbone of the cable company.  Your speed is dependant on how big their transit port is to the big, bad, porno-filled Internet, in other words all the users of the cable company. All the users of DSL in your area all run to the same DSLAM.  Basically a multi-plexer for your DSL.  So if that particular CO (central office) where the DSLAM is has a 45 meg connection to the actual IP network of Verizon (and by extension the actual Internet), then the congestion model works just as well as the freeway does at rush hour.  As soon as Verizon sells a bunch of DSL connections in your area, BOHICA. G23c Sorry no sell I have been on lots of shared networks that slowed down as traffic increased. And yes a Ethernet network will absolutely slow down when everyone is there and you exceed throughput capabilities of the network, seen it happen many times. You ever heard of sub-netting, one of the primary reasons to sub-net a local network is to increases efficiency and cut down on slow downs due to increased traffic. There is absolutely NO DOUBT my cable connection speed dropped EVERY weekday afternoon between 3:00 and 11:00 PM. The reason is obivious. If your cable company puts to many people on one loop YOU WILL SEE SLOW DOWNS.

Ahem.

In a non-switched environment (rare nowadays) when everybody is on the same segment, network traffic becomes an issue due to collisions and the fact that every packet goes out every port of a hub. Fast forward to today. In switched environments, packets are only sent to ports on which the destination layer 2 address is found. Every machine has a dedicated 100 MBit connection to the switch. Collisions are eliminated. If the switch is a pile of crap, switch utilization MIGHT be an issue, but this is the fault of the switch and is NOT a problem solved by subnetting at layer 3. If there are 3000 machines clogging a 100 Mbit link to a shared departmental server, this isn't solved by subnetting either. If you have 24 active users on a switch and are experiencing slowdowns, how does breaking the ports into VLANS and indtroducing routing at layer 3 fix this? If anything, you've just added another hop for no good reason. Subnetting places devices into separate broadcast domains. You need ALOT of machines in the same broadcast domain to create a traffic problem. Decent modern switches can limit broadcast storms, so subnetting for "efficency" is kind of silly. Taking it a step further, implementing QoS gives priority to business critical apps, so subnetting becomes even less a reason for "efficency".

There are alot of reasons that networks are divided at layer 3, and the ones you listed are not the primary ones.