I Tested EVERY Ethernet Cable... You WON'T Believe the Results!

[@landpet]

Product Links: cat8 ethernet cable: amzn.to/4fiFTqs cat7 ethernet cable: amzn.to/3QIm3dT cat6 ethernet cable (similar to one in video): amzn.to/3ZVNeI6 cat5e ethernet cable (similar to one in video): amzn.to/41Ea6go NETGEAR 12-Port 10G Plus Switch (XS512EM): amzn.to/3OH7o22 NETGEAR ORBI 970 Series (2 Pack): amzn.to/3S68oja As an Amazon Associate I earn from qualifying purchases

[@alexanderlakey8483]

Sorry, at the lengths you are testing it makes no difference. The whole point is how crosstalk and interference impacts longer cables..

[@landpet]

I actually wanted to know and some have asked as well. I genuinely didn’t think it would hit these speeds. The interesting question would.m be at what distance it makes a difference

[@AndrewBatiuk]

@@landpetsounds like content for a follow up video!

[@hariranormal5584]

right, short distances hardly matter

[@alexanderlakey8483]

​@@landpettry 25 metres.. that should show you a difference and 25m patch cables are available.

[@minixtvbox]

Yes

[@Brewdog2001]

If you want to show how these cables really get effected, get longer cables and an RF source that is near a harmonic of the wavelength that the ethernet is using and watch how much that effects your speeds per cable. That will really show that its about crosstalk, and interference across the length of a cable. Remember a random length of wire is basically an entenna.

[@rogerphelps9939]

There is no such thing as "the wavelengh the ethernet is using." Ethernet is wideband.

[@marksapollo]

Cat 6A is all anyone needs. Cat 7 is a fake standard as the official spec of it doesn't use RJ45, If you need Cat 8 speeds you should really be looking at fibre.

[@landpet]

cat6a does deliver 10Gbps speeds, should be golden with that.

[@x12flare73]

Also those cat 7/8 cables you see on Amazon are usually just badly made Cat5e cables using CCA so it’s best to avoid those.

[@defencebangladesh4068]

Yep

[@ram64man]

i disagree, shielded cat 7 is superior, to cat 6E, sure there is the limitation of 150 ft but that is fine for domestic, further more with domestic electrical 6 and 6e do get affected especially on + 6amp - 32 amp circuits, i never buy premade i agree that you don't get what your paying for, and opt for cat 8 keystones. its great for 2.5 g , the issue with fiber adapters get super hot and unsuitable for domestic, its not as flexible for install and cable is super expensive, plus you would need a break out router in room that would be compatible usally taking it to 2,5gbit anyway, the only time i run fiber is if in a property that needs to go external annex or greater than 1900ft or usually multi apartment dwelling scenario, with 4- 10gb Ethernet circuit incoming, to central 10gb central lev 3 switch/router.

[@marksapollo]

@ firstly it’s cat 6’A’, secondly cat 7 with RJ45 is not the cat 7 spec, it is a fake spec that doesn’t actually exist and was made up just to shift more cables, but cat 6A is better as it’s the set spec that you can buy. Cat 7 is meant to be used with TERA or GG45 connectors, GG45 is compatible with RJ45 equipment though, I believe it’s to do with the shielding in the cable and the grounding required to make it work properly. Cat 7 also was never ratified by the IEEE as it’s proprietary, if you don’t use the proper connectors and grounding, all that shielding is wasted and serves no purpose.

[@TheUweRoss]

Way back in 1996 when I built my house, 10BASE-T and CAT-5 cabling were the hot thing, so I installed CAT-5 all thought the place before the drywall went up. Some of the runs are well over 50 meters. Today, it all runs gigabit just fine and I really can't see needing more than that.

[@landpet]

Nice!

[@NotHimJim]

I have several cat 5E long runs in the house and they can all achieve what my equipment can handle (2.5Gbps).

[@FiveOhWill]

Same for me. Just one long run though.

[@Solkre82]

2.5Gb is the savior of 5E

[@defencebangladesh4068]

Same here. They also send data farther than cat6

[@landpet]

Oh nice! How long is your longest run?

[@defencebangladesh4068]

@@landpet mines are 100m

[@NotHimJim]

I'd be curious to see at what distance(s), the cable matters.

[@landpet]

That would be interesting to know since all of these are short cables

[@IlanEliyahu]

​@@landpet can you please do another video with much longer cables?

[@RainKvks]

YES! I NEED TO KNOW

[@Matlockization]

@@landpet Perhaps you could redo the test with longer length cables, provided of course you have already done your research beforehand.

[@BinaryBlueBull]

@@Matlockization Good idea indeed, I would be interested to see this as well. And then perhaps he can gradually increase the length each time, until it starts impacting the speed? I would be veeeery curious to see at what distance the CAT5 starts to significantly drop off

[@jylfarm1964]

-- You should not purchase CAT 7 -- that not iso-standard... only industries standard. However, should do 10 Gbits/s just fine… They are typically equivalent to CAT 6a. -- No reason to purchase CAT 8 -- unless you have 40 Gbits/s equipment. At 10 Gbits/s, CAT 6 should work until 55 meters... and 6a to 100 Meters At 10 Gbits/s CAT 5e should be good for 35 to 45 meters. Digital cable either work or they don't. Should not change speed unless they downgrade to 1 Gbits/s. Regarding CAT 5, you need very old cable as manufacturer still produce them but with the CAT 5e specification. Essentially, newer CAT 5 and CAT5e are the same thing for at least 5 years.

[@zorin1us]

What you need to be doing is monitoring for dropped packets. I don't know if the router or the Ethernet port will renegotiate down to a lower speed if it encounters to many dropped packets. Remember that dropped packets will be retried which should show a slower transfer rate based on the size of the packet you are sending. Dropping jumbo frames will have more of an impact. You should monitor your dropped frames to see how well each cable is doing. A speed test alone does not show the whole picture.

[@landpet]

That’s also a great point

[@michaelbeckerman7532]

Correct. Let's not forget here that the whole reason why we even have higher categories of cables in the first place it to be able to buy cables that are CERTIFIED to provide a certain level of data throughput WITHOUT an abundance of packet loss. This is far more about data and signal integrity than it is about speed and throughput. What a cable can do and what it can do reliably are often two VERY different things. Sending packets at high speed does you no good at all if you can't actually rely on the fact that what was sent was actually received properly down at the other end.

[@whophd]

Yep this is a great point. Ethernet - and the protocols that usually sit on it - handle a few dropped packets fine here and there. It’s not the same story for serial interfaces like Thunderbolt and HDMI, that need to send a STEADY MINIMUM stream of data at all times, and immediately get glitches visible if any data drops. Very different purposes. So if you abuse ethernet like this, it tends to forgive you. Remarkable, honestly. It’s a whole story, how this is a very loved technology that has served many people and their careers without as much pain as any of the alternatives.

[@zorin1us]

@@michaelbeckerman7532 Maybe looking at dropped packets is best, maybe looking at errors instead. Wifi has a lot of dropped packets. No one really reviews dropped/error packets but just look at how fast things are.

[@alb12345672]

@@whophd The power industry streams all their control and monitoring thru ethernet cables. Dropped packets are very serious!

[@michaelmanus7765]

I am unclear why you bother with very short cables. Most people don't have all their networks on "very short cables." Try different longer cable lengths. Also, do packet testing.

[@landpet]

I may have to pickup longer cables to try

[@geomatrix5452]

It's "Cheaper" hahahahahaha!

[@vadnegru]

if you buy a fancy wifi router all you get is a slim cat5e. This test shows that it`s fine

[@MichaelSmith-mw8ed]

I have not noticed anyone talk about the ability to keep interference out at those speeds. The better cables tend to reject noise better than the cheaper cables. I have substituted CAT6A in noisy environments and it made a difference. Obviously, fiber trumps all these as it basically has no interference, and the lengths can be in miles. As fiber has gotten cheaper, I have used more fiber in client environments and the armored stuff is great for pulling through existing chase ways and conduit. Short cables will do fine for almost all speeds up to 10g and this was a great demonstration! Longer distances do make a difference as I have run into it firsthand. Again, great work.

[@wildmanjeff42]

I went to mainly 10G fiber between my servers after lightning took out 2 of my switches and a camera, Access Point. Just to make sure.

[@michaelbeckerman7532]

@@wildmanjeff42 That's actually really smart as no fiber optic cable can ever pass a surge or a spike the way a copper cable can.

[@wildmanjeff42]

@@michaelbeckerman7532 I run everything on UPS, and surge supressors, and know the lighting came in on low voltage landscape lighting into power and jumped onto my network.

[@The.Orchard]

CAT5 STP will keep interference out better than CAT6 UTP. But it does nothing to stop interference between pairs.

[@Matlockization]

What are these 'long distances' before they make a difference ?

[@fr33for3]

Test 100' cables, this is where the differences begin to really show.

[@AndrewLumsden]

100' is only 30 metres. He should be testing 100 metre cables. - 328 feet.

[@melgross]

@@AndrewLumsdenwell, that’s an extreme. If there’s a reduction at 30 meters the loss at 100 meters can be calculated.

[@trevordennis]

@@AndrewLumsden100' would be a typical run from your desk to the office wiring closet where the switch is.

[@francocastilloAR]

@@melgross It doesn't have to be linear.

[@mikespangler98]

I have a 120 ft run to a detached garage. It's underground, I don't know if that makes it better or worse. It's also possible the conduit is full of water. Gigabit is still working though.

[@coldshot1611]

Cat 7 is not a recognized standard it was more of a marketing deploy. Cat 6a is better than cat 7.

[@landpet]

Interesting, thanks for that info

[@brahmincalgary9239]

Please try this. Do another test. 50 feet long Cat5 or Cat5e versus 50 feet long Cat 8. Use all the same hardware and software you used for this video. See if there's a speed difference.

[@landpet]

A few more videos coming out on this topic with longer cables. I'll try to put one out in the next day or two.

[@Whosever]

Cat 6/6a it's more than enough for end user, it can even handle 10g up to 100m, if you need more than that (longer distances or faster bit rate) then go fiber

[@Matlockization]

How do you hook fibre to your desktop ?

[@zakuso5141]

@@Matlockization You can install a PCIe Fiber Network Card , or connect a USB-C / Thunderbolt Fiber Optic Network Adaptor

[@robertct06]

4:08 “Make the video as short as possible”. Love you. So many people just try to make it as long as possible and I can’t stand it lol

[@landpet]

This one was almost real time. My other videos, I try hard to cut it down to as short as possible. I just wanted to show everything being connected so people would see I’m actually using the cables

[@robertct06]

@ I meant to put that in parentheses to show I was quoting you lol. The video was awesome

[@bigjohn2811]

If you have CAT 5e in your house, it will probably 10 gb without a problem. If installing new cable, then use cat 6. It doesn't cost much more than 5e, $25 to $50 more for 1,000 foot bulk roll. Use 6a if need to guarantee 10gb connection in a noisy environment or very long cable runs.

[@PatrickDKing]

I'd be willing to bet that all the cables even at greater lengths will perform the same. In the end it's just the element copper. I'd like to see all categories at 100 feet length being tested. Also, that's awesome that you can get 5 gig internet where you are. We get fake gig here with Cumcast, allegedly 1.2 gig down but only 25 mbps up...purely criminal.

[@ContraVsGigi]

In perfect conditions, maybe. But add some used electric wires around and then count the dropped packages. Expensive cables have extra features, for instance a metal mesh (or similar) protecting the signal from external factors.

[@chefmike8888]

Well that’s exactly what I expected. The reason for thickness is shielding. CAT 5 has little or no shield. 5 E has outer shield, CAT 6 depending on manufacture has 2 shields, CAT 6E does not exist. CAT 7 has each pair shielded as well as outer shield. The only reason a home should run 7 is if it’s new construction and you want to future proof it. But it’s going to be years before all equipment is at the CAT 7 standard. Again for (basic) home use 5E or 6 is more than enough. CAT 8 is a gimmick. In the 15 or so years it will take to actually be at the point where 8 is implemented the whole CATegory may very well be old technology and a new technology will be in place. As these test show the number on the cable makes no difference simply because of the ISP is the one setting the speeds. All the shielding does it help protect from interference the cable may encounter in its path due to electrical devices or other data cables run in the same raceway. All the conductors within each cable is the same gauge within maybe a gauge or 2 . What was your switch being fed with? Try feeding the switch with the different wires. Or go check what your ISP is feeding the router with. If it’s 5e then that’s the max you should buy. Because feeding a 7 with a 5 will still only transfer 5 .

[@landpet]

Everything is cat7 or cat8 at my place, mostly cat7 though. I’m going to run some length tests on the cat5e and cat6 cables

[@plop31]

CLICK BAIT! !!!!!!!!!!!!!!!! I wanted a test on the length of use in an ACTUAL HOME, not a one meter cable..... With a minimum length of 15 meters, real use is closer to 20 to 30 meters in length. i'm pretty sure a old CAT 4 cable can do the trick on 50cm length

[@landpet]

I'm ordering longer cables, there will be another variant with much longer cables. Similar tests

[@jaakanshorter]

You will never hit 10G due to overhead.

[@บรรพตหล่อทรงธรรม-ฟ4ท]

Cat 6A 10G

[@barryc573]

This test is not valid as the higher cabling standards have extra shielding for interference and crosstalk, and guarantee the speeds at 100m (330 ft). You can easily get 10Gb on cat5 over 2-3m

[@Omar-hu5qt]

You’re thinking of the old standards where cat 5a was 1 Gbps. New standards allow 2.5 and 5 Gbps on cat 5.

[@mikezappulla4092]

It still did much faster than that.

[@landpet]

When did it change?

[@UniPer85]

The standard never changed as far as I know. The default test method is running the cable for 100 meters at maximum speed, which it should maintain. Cat5e cable is tested at 100Mhz for 100 meters and guarantees a speed of 1Gbps. With the condition that you use a copper cable and not an aluminum core with a copper jacket of some kind. Nowadays some manufacturers offer Cat5e cables tested at 350Mhz which can deliver up to 10Gbps. The only problem is that they rarely tell you that it's only up to 10 meters distance and perhaps not fully stable. Therefore it cannot be marked as a certified speed. So...bottom line: Cat5e cable is certified for 1Gbps Cat6 is also certified for 1Gbps (at 100 meters) but can perform at 10Gbps on 55 meters. If you want a full 10Gbps for 100 meters, you should use Cat6a.

[@Omar-hu5qt]

@@UniPer85 IEEE 802.3bz, NBASE-T and MGBASE-T standards released in 2016. en.wikipedia.org/wiki/2.5GBASE-T_and_5GBASE-T Before this cat5e was only rated for 1 Gbps. Now it's also officially rated for 2.5 and 5 Gbps.

[@Omar-hu5qt]

@@mikezappulla4092 Ya but only for short runs and low interference.

[@KingdaToro]

Cat5 is rated for gigabit, not 100 megabit. The gigabit spec was developed specifically to max it out, because it was the best cable available at the time. Cat5e came out a few years after gigabit. The actual rated speeds at 100 meters (important!) are 1 gig for Cat5, 2.5 gig for Cat5e, 5 gig for Cat6, and 10 gig for Cat6a. Cat7 and Cat8 should not even be considered, they're not real Ethernet cable. And, besides, for speeds of 10 gig and up, fiber is actually cheaper both in terms of hardware cost and power consumption.

[@halduncincioglu6911]

Since the length of the cables is short, it is natural not to see a difference in speeds. This test only makes sense with a minimum of 300 ft cable.

[@landpet]

I ordered various length cables up to 300ft (91m) for cat5e and cat6. New videos coming soon.

[@adventureswithdavedave]

Good start, but as others have mentioned, get another 90 meters and run them in your attic to a POE device and do more testing. And then compare all that to fiber. Make sure to include power consumption. The end result will be, just use fiber.

[@landpet]

I'm sure fiber is better. I'm going to run the cat5e at varying lengths, just got the cables today. I will also do one for cat6 soon

[@nathanwieling7943]

Cat 5 cables have little to no twist in the wiring, Cat 5e has loosely twisted wiring (helps reduce crosstalk) and Cat 6 cables have even more twist in the wiring. Shielded ethernet cables reduce external interference. The speeds the cables are rated has more to do with distance and crosstalk degradation with some extra buffer so the speed can be guaranteed. If you want to get serious, get 50m or 100m Cat 5, Cat 5e and Cat 6 cables to see how they handle when you unspool them and run them near some power wiring in your home, etc. At that point you should start to see the limitations of cable design and quality of construction.

[@landpet]

I’m going to do a new video for cat5e and cat6 for varying lengths up to 300ft (91m)

[@jfiosi]

Glad to see the test, but testing longer cables would be been more informative and helpful in buying decisions. I have all the CAT speeds you showed. I paid more for my CAT 8 cables. Now I wonder if I should have stuck with my CAT 5. I am going to switch cables to see if I see a speed difference. Can you do a test with longer cables, or is that cost-prohibitive for you?

[@landpet]

I bought the cables, I will be doing more tests. A few more videos coming out soon. Cat5e is first as I got all the cables today. Cat6 is next

[@dexterjsullen]

Another aspect to remember file size (duration) often times they drop off after continuous use, do a file transfer of 10gb files

[@Tomcat2_kanal]

Actually you are getting 10Gbits on the physical layer but for the application data transfer speed you will always see less as there are other reduntant data like Ethernet frame and IP packet headers transferred, control data, CRC, etc...

[@landpet]

Makes sense. I’m going to redo these tests with longer cables for cat5e and cat6 as well. From various lengths up to 300ft (91m)

[@EugWanker]

To be honest, I'm not sure how helpful these tests are, since the cables are so short. I ran Gigabit over CAT3 and it worked fine, but again, it was a short cable (less than 2 m). However, I would never recommend anyone run Gigabit on CAT3, unless that's all they had in the walls and are aware it might have some problems over longer runs. That said, I've since upgraded to a 2.5 Gbps home network, and I've found some of my CAT5e patch cables are problematic. I noticed a couple of runs acting flaky at 2.5 Gbps. Swapping out the CAT5e patch cable for a CAT6 cable solved the problem. The first example of this was a long CAT5e patch cable (solid copper) that was connected to a CAT5e run (solid copper, I'm guessing around 40-50 feet). It negotiated fine at 2.5 Gbps and seemed to work fine at that speed the first day without dropped packets, but I noticed after a few days that it'd would periodically drop down to 1 Gbps. I changed the 30 foot CAT5e patch cable to a 25 foot CAT6 patch cable and the problem went away. The other thing I noticed is that with a couple of other runs with in-line CAT6 or CAT5e couplers, they worked fine at 1 Gbps, but would not run at 2.5 Gbps at all. I removed the couplers and they work perfectly at 2.5 Gbps. I know you're not supposed to use couplers, but since they worked for my Gigabit network, I had previously been too lazy to remove them. tl;dr: It's well known that short CAT5e cables will work for 10 GbE, but longer cables will be problematic, especially if it's not an optimal install. If you're trying to run 10 GbE, or perhaps even 2.5 GbE, then save yourself some headache and when possible, try not to re-use your old CAT5e patch cables, particularly if they are long. P.S. You probably shouldn't buy so-called CAT7 and CAT8 cables. Most of the time you'd be better off buying legit CAT6 or CAT6a cables.

[@SandyFunnies]

Years ago I was told the only difference between a CAT5 and CAT5e cable was the level of testing done by the manufacturer. The materials used was identical and the only difference was the test run on lots from the manufacturing process. I have noticed that CAT6 cables use larger copper wires than CAT5e so I'm guessing they are using larger gauge wires. Your test was great and confirmed what I was told years ago. I liked that you kept the test simple with the same hardware, software, and cable lengths similar. Thanks.

[@BB..........]

Do a proper test with 90m/295 ft of cable for each one.

[@STLJonny]

Redo all test but use 25 to 30 foot versions of each cable. That will be a real test. You have to understand that those cables are rated for that category at a minimum. It’s certifies that they are able to achieve that speed as a minimum. Maximum will depend on a lot of things, as well as two cat 6E rated cables can test out totally different.

[@smert_rashistskiy_pederacii]

7-8 meters is not the change either. I mean, what's the point? Recently I wanted to replace one of my cables in house and I measured the approximate length of it to be about 15m minimum. My friend earlier used cheap version of CAT5 cable to reach his neighbors on the distance of about 100m and it also worked perfectly but for 100 Mbits only. My 15m CAT5 cable works just fine with 1 Gbit on that 15m as well.

[@parawm4585]

Indeed, please test much longer distances. We have to know :) I just tested 7 long and medium Cat 5e cables I had lying around in series with 6 of those small adapters in between. Total length 58m and they were all rolled up still, for worst case scenario. Still did 2.5 gbps, can't test faster..... Really like to know at what length even cat 5e sees any degradation below 10 gbps.

[@07BSPtC]

You won’t get to a perfect 10 there is overhead. 9.8 is about it.

[@landpet]

Makes sense

[@AnupomAG]

Basically there was almost no difference in speed orl latency and the reason is distance. Regardless I was really surprised that CAT5 could reach 10 Gig, heck I never thought even CAT6 could reach that speed. 9:00 Yes you been repeating that word but still 😮 why is that even possible? Thanks for the lesson ❤

[@michaelbeckerman7532]

I think the key issue here isn't actually speed at all, but really reliability and data integrity. These cables are all still just made with copper wire essentially (lower-end ones being made with CCA, Copper Coated Aluminum). They can all push the same amount of speed/throughput. The difference lies in which ones are actually CERTIFIED to be able to do what speeds reliably with a high degree of data integrity. I think the main differences between some of these cables possibly comes in the quality of the copper that is used in them as that can have an impact on signal quality and therefore data integrity. And yes, twists per foot and cable shielding certainly play a role here as well Just like in the world of HDMI cables, as the quality of the cable goes up, the grading and quality of the metals used inside of them typically goes up as well - right along with the price. The idea there is to always try and lower the impurities inside of the cable (Long-grain copper, OFC cable, OCC cable, etc.) so as to reduce the likelihood that the signal will be disrupted or disturbed as it is passing down the wire.

[@johndoughto]

running cat5e from 2002 that i cabled in my house when built. running 10gbe throughout. sfp modules, dac cables and compute hardware are my bottle necks for 10gbe - not my cables. usually run whatever patch cord i like for diameter and color, and they all work. was doing usoc before "cat" was a thing. obviously i don't run out to 100m to test limits, but my 3000sqft house cabling works. don't get caught up in the CAT.

[@บรรพตหล่อทรงธรรม-ฟ4ท]

Cat 5E 2Gbps

[@thomasmaughan4798]

"I Tested EVERY Ethernet Cable." I test every cable I make, but I do not test EVERY ethernet cable; there's millions of them out there.

[@landpet]

I’m testing more with varying lengths on cat5e followed with another video for cat6

[@Archangel-v5y]

As people have pointed out that on short distances you're ok. The cables aren't 'thinner' IIRC the wires in the cables are the same thickness. What you're seeing is more shielding which is what will let you go longer distances. If you take cables at 20-50 ft in length and have a 'noisy' environment you'll see the differences.

[@Richard_GIS]

0:37 Nice wallpaper from Bing, Salzburg Austria, love it! Greetings from Austria ;-)

[@landpet]

Very nice place!

[@whismerhillgaming]

several things 1 - the lenght is too small 2 - speedtest OMG really ? you realize you are dependent upon the flutuations of the internet with that ? which severely limits the accuracy of the results So no just no, need Local testing not internet testing 3 - testing for dropped packets, dropped packets are actually a much more interesting metric since those may hamper for example your multiplayer gaming sessions by making your latency irregular 4 - not sure but different switches might also influence the results

[@landpet]

I ordered various length cables up to 300ft (91m) for cat5e and cat6. New videos coming soon.

[@Winnetou17]

on 2 - OMG have you not seen the video ? It's local and very clearly so.... maybe watch the video before complaining and making a fool or yourself

[@dave24-73]

I have Cat5e in my house, due to age, and I get full speed, just goes to show in an average home, you may not need the extra spec. In industry where you often have lots of cables and long runs the higher spec cable is likely to have more losses, and thus become more important.

[@TheRealMrBlackCat]

I guess most viewers don't work around server rooms where this information is very important... I learned this in the late 90's switching off Coax (10Mbps) to 10/100 and some 1Gb trunk runs. I stock 5e patch cables in lengths from 4 inches to 4ft in every color basically... it is always enjoyable to demonstrate that "high speed" cables in this setting are mostly irrelevant.

[@todoralexandrov1]

You'll never see any difference in speed with a short patch cable.

[@landpet]

I’m going to do another video with much longer cables. Ordering them shortly.

[@roberts.3712]

This test was a waste of time.

[@olafschermann1592]

Curious why RJ45 makes no problem for everything above 2.5G. Back in the years they discussed about a new connector because of the critical distances between contacts.

[@jasontemple4407]

There are a few things to consider. Distance is the first thing as mentioned many times in the comments. The next thing to take into consideration is Line rate. The cables are short enough to negotiate at the 10g line rate. You had 1 cable under perform. The end points being used are through one switch. An ideal test case would be PC 1 to switch 1, switch 1 to switch 2 using the different cables, then switch 2 to PC 2. Then you could get error information on both switch ports where you are swapping out cables. On the under performing cable I would be curious to see receive errors on both devices.

[@timothyhoadley2432]

As I understand it, the categories of the cables designated by the numbers references the speed at which they are supposed to be used for is an FCC criteria. The higher the number, the less RF radiation leaving the cables which adds to the radio frequency interference in an area.

[@francocastilloAR]

9:16 Not true. They are rated to operate FROM that speed.

[@ericv4720]

Hello. Would you able to repeat this test on a 100 ft ethernet xable for cath 5 ro 8? Ty. Looking forward to your next post. Excellent job!!!

[@landpet]

I'm trying cat5e and cat6 at various lengths up to 300ft (100ft will be part of the testing). I may do some additional as well.

[@Dysl3xicDog]

The speeds you're getting are over a VERY short distance the spec of the cable is for 100 meters you would likely have a lot more issues if you had full distance runs to run these tests with.

[@LPChip]

I'm fairly sure the limitation for not getting the full 10gig is the switch. As you download, you also need to send confirmation packages, so you consume upload bandwidth as well, and I think, the 10gig throughput is both up and down streams combined. So while technically the download could reach a full 10 gig and the upload as well, it would only do so on UDP where it doesn't send the confirmation packages. I also think that is why the upload on the first try with cat7 was much lower. The upload was going up and down the whole time.

[@MalamIbnMalam]

I want to get CAT 6A wired all throughout my house and get outlets in practically every room. Would they have to break down walls for that?

[@iandron7119]

I’d always suspected this was the case and nice to have it confirmed. Nice video!

[@FanfanPO]

Thanks for the test! At first, I thought you should standardize the length of each cable to keep the variables consistent. But in the end, the relatively long and old CAT5 cable still performed really well. I’ve always known that as long as the cable is short enough, signal loss shouldn’t be significant, and cables can often exceed their rated specs. Still, this result exceeded my expectations. It seems these standards are designed for long distances and harsh environments, while for typical home use, we might not need to worry about them as much.

[@lucsteffens]

I agree, i'm amazed to see the only diff between the cat versions is just 'noice cancelation' and is actually irrelevant on shorter distances. Thx for the test!

[@joeythedime1838]

How about running this test with an old school CAT3 cable?

[@michaelbeckerman7532]

For that matter, why not run it on Flat Satin?

[@landpet]

I’m going to rerun this test with various length cables for cat5e and cat6. We can see how big of a difference distance will make.

[@SteelBlueVision]

Nah, let's see four pairs of untwisted lamp cord. It doesn't even have to be copper.

[@joeythedime1838]

@@SteelBlueVision Us old guys remember Cat 1 being used by Farallon's PhoneNet which implemented Apple's LocalTalk over Cat 1 at a blazing data rate of 230.4 kbit/s.

[@thomasmaughan4798]

@@michaelbeckerman7532 "why not run it on Flat Satin?" Flat cables work great on short patches and also easy to terminate.

[@Proton_Decay]

100% this depends on the amount of interference and adjacent wiring. I live in a rental where the ONT is in the property manager's data closet, and I have patch ports around the house but only choose ONE for the WAN to go to. I got the tech to patch two rooms together in the data room, so there's a LONG Cat5 run, out of the townhouse, to a data closet, and back in, and I am able to maintain a 10GbE backhaul between the upstairs and downstairs no problem!

[@mymymy9kh]

Test 100m / 200m and 300m lengths and then you’ll have a better test

[@SkaggsFamily]

You can even use Cat3 or uncategorized "twisted pair" telephone cable and get similar results for even long runs if there are no other interference sources, like a dozen other cables running parallel, or excess length coiled up. I built my house for networking back in the 90's with two runs of Cat5 (one for phone, one for ethernet) and two RG6U to every room. They only cross electrical romex as 90-degrees, never along side or on the same stud. My hardware has upgraded over the years, and I'm running 10GiB to a NAS for real-time video editing without any issues.

[@landpet]

Thanks for the info, I’m going to run more tests with longer cables to see what happens.

[@smert_rashistskiy_pederacii]

Good to know that CAT5 is still in business, no matter what. And why not? 4 pairs of cable is still 4 pairs. Not 2 like long ago for 10-100 Mbps connections.

[@alexdsr1]

Finally it will be cat of 9 tails. 😊😎

[@MadawaskaObservatory]

what about range?

[@landpet]

I'll be doing a distance test for cat5e from short cables all the way to 300ft (91m) I'll do the same for cat6 as well

[@indiemichael]

Ethernet can go up to 800Gbps. You’re confusing Ethernet with “Ethernet over twisted pair” or Base-T, Base-Ten

[@thomasmaughan4798]

"You’re confusing Ethernet with..." Oh so much misinformation on this page.

[@CrazyAndy1983]

You can never achieve exactly 10 Gb/s on your speedtest because of protocol overhead. The 10 Gb/s is the line speed not actual speed up on the OSI Layer.

[@stoptellingmewhattowrite]

There was decent amount of CAT3 testing done (on multiple forums) and I recall one gigabit was doable usually to at least 100 feet (did at times depend on cable quality). So this does not surprise me that much, especially with very small length of cables. Longer lengths should be explored, but thank you for the video!

[@davidhalliday7776]

1) CAT5e was a change in the testing standard and not necessarily in the cable manufacture. 2) Most (except the crappiest) CAT5 will test fine as CAT5e 3) 10GBASE-T was designed with CAT5 in mind 4) CAT5e is good for 10GB/s up to 45m (about 148ft) 5) As CAT5 and 5e are about the same, expecting 10Gb/s speeds up to 150ft on CAT5 is not too crazy. There is also STP cable (Shielded), which can improve matters, BUT every connection, cable, and item in the path must be STP. Put one UTP item in the path, and the whole path may as well be UTP.

[@4givenoxygenbarez]

at what distance are cat 6 and below done?

[@landpet]

That's a good question, I don't know in practice. I may have to buy the cables to try it

[@aallien]

I’ve definitely swapped out a cat5 cable at a customer and drastically improved speeds. It’s usually those old blue ones that came with the WRT54G linksys routers. Folks plug that into their modem and swap out the router along the way and speeds change coming in but the customer doesn’t realize the 3ft cable is the bottleneck. I believe the quality of cable you’re using must be the cause of good speeds regardless. Puzzling results based off my own experience with the link speed I’ve seen switches auto negotiate to with a cat5 cable.

[@DanielMartinez81]

Based on my experience, if you have less than 2m or 3m cable then no matter which category is the cable, they will handle the full bandwith. But for longer lenghts you will probably prefere to select high categories if you want to ensure a proper bandwith. Anyway nice video!

[@andrebrait]

I have full 10GbE running on several ~10m long runs of Cat 5E cable. I have never observed any speed or package drops.

[@landpet]

Nice!

[@บรรพตหล่อทรงธรรม-ฟ4ท]

Cat 6A 10G

[@grasstreefarmer]

Yeah its as expected. There is no limit to what speed a cable will do. The standards just specify a minimum over a certain length. All the cables are just 4 pairs of wires so they will all work the same over short distances. Its the distance that makes the difference. The longer the cable the higher the resistance of each wire and the bigger the antenna is for interference. That's where the tighter twisting of the wire pairs, pre formed internal separators and shielding of the higher standards comes into play. Pretty much all 'cat7' and 'cat8' cables that you can buy from normal retailers don't truly meet the standards they are sold as. That's why professional cable installers still make their own cables and its expensive even with real cat6a. The retail patch cables work but they use thin gauge wire and I would be hesitant to use them for POE applications. The flat and braided cables are the worst offenders. They use wires as small as 30awg so their resistance gets really high over any real distance like 10+ meters.

[@landpet]

Ive already ordered various length cat5e cables up to 300 ft and the same for cat6 as well. I will be redoing these tests with various length cables when it comes in.

[@thomasmaughan4798]

"There is no limit to what speed a cable will do." More misinformation :-) But who cares.

[@802Garage]

It does make some sense given the cables are completely passive and as long as the wire passes the signal properly you get the speed. As you said, distance is key, but I'd be interested to see how much it really matters for any given distance.

[@thomasmaughan4798]

"I'd be interested to see how much it really matters for any given distance." It matters a lot and for several reasons. Interference and crosstalk is a biggie, but so is smearing and group delay. The bits start out nice and square but by 100 meters, they are squishy and round; difficult to know where exactly is the edge of the bit.

[@802Garage]

​@@thomasmaughan4798 I totally understand all that. I'm saying I'd like to see the testing is all. There are obviously a lot of factors that can come into play. :)

[@karnic-f1x]

The networking circuits will test the quality of the cable when you plug it in. It will connect at highest possible standard speed 10G, 2,5G, 1G, 100mbit or 10mbit, full or half duplex, not anything in between

[@landpet]

Going to run various length tests up to 300 ft with cat5e and cat6 to see if that makes a difference

[@karnic-f1x]

@@landpet Interesting. There is a dynamic tuning of the signal depending on the length of the cable (delay). Sometimes you can read out the length of the plugged in cable directly from the NIC driver

[@imark7777777]

I was cleaning out the closet where I work and found some cat 3 and cat 5 cables and I was thinking of doing something like this but I can only go up to 2.5 gig. Just mostly the test functionality. Trying to decide whether I should return them to optional service or if I should make everybody's life easier by making them disappear.

[@landpet]

Your choice. I’m going to retweet cat5e at varying lengths also do the same for cat6 up to 300ft (91m)

[@homemark22]

I only have a cat meow!

[@landpet]

lol

[@Swecan76]

Why are you testing with short cables barely even 1 meter in length? Kind of defeats the purpose of the testing. Do a 20ft cable instead and see if any drops start to happen.

[@IntangirVoluntaryist]

This is good, my house was built with cat 5 built into the walls, glad to see it's pretty much as good

[@woolybooly6859]

Very Helpful and eye-opening results with older cat 6 cable. I noticed you are in the LA area (I am too). I also select the CTCSCI server when using Speedtest. Question: I have an M1 Mac Mini, 16GB 256hd. I also just upgraded to Frontier 2Gb service. I know the ethernet is only 1Gb. Can I get 2Gb speed with some kind of adapter on the USB-C? (less than 25ft to the Eero Pro 6E that Frontier provides). Thanks for the help

[@landpet]

Yes you can Is This the FASTEST USB-C to Ethernet Adapter? WAVLINK 5Gbps Tested kzbin.info/www/bejne/e3jQfoNtbc53ipI

[@WonPXL]

It’s a nice idea to test cables - but you need to be testing rolls of 100 metres. The cables are rated at their install lengths.

[@ehsnils]

You miss the Cat 3 cable.

[@landpet]

Cat5 was the oldest here

[@ChinchillaBONK]

imagine after all that he realised his wifi is turned on. 🤣

[@RolandGustafsson]

My house was wired CAT3 in the 90s and some of the ports can handle gigabit but the far end of the house is only 100Mb. Your video made my realize why the far end is slow.

[@landpet]

I ordered various length cables up to 300ft (91m) for cat5e and cat6. New videos coming soon.

[@bastel]

Very interesting! It would also be interesting to know if Cat 5 cables interfere with each other when many are stuck next to each other in a switch or patch panel.

[@thomasmaughan4798]

"It would also be interesting to know if Cat 5 cables interfere with each other when many are stuck next to each other in a switch or patch panel." Yes, they do, but not significantly.

[@Snerdles]

It would be interesting to test 5m, 15m, 25m, 50m, and 100m to see how each hold up to different distances.

[@landpet]

I am ordering the cables at different lengths, I will be doing that for cat5E and cat6

[@Xico-x2l]

2:05 *9791.1 Mbps !?!?* Meanwhile, my speed test: 5 Mbps / 1 Mbps

[@DelticEngine]

Interesting video and thanks for making it. On paper the maximum length of a patch cable is 5 metres, so this would be a good length to test. Patch cables use stranded conductors and do not work as well at higher speeds which is why installation cable is all solid core. What is worth noting from what you have done here isn't so much the achieved speeds but the increase in ping time and jitter as the specification of the cable goes down.

[@Winnetou17]

Wow, the title clickbait is strong in this one: "EVERY cable" - just not the Cat6A, literally the exact one that everyone should buy now. I do appreciate the activity in the comments and the desire to test further and better. Good luck on that!

[@landpet]

I did mention all the cables in the thumbnail that I will be testing. I’ve also ordered various length cables for cat5e and cat6. I may do cat6 also but I feel like that one will do the full 10Gbps even at longer distances.

[@Winnetou17]

@@landpet Fair enough, and thanks for the response. The pedantic in me is still not fully satisfied. Can you add an asterisc ? EVERY* cable blah blah. It's not just that one is missing (and besides the thumbnail and the description, it is acknowledged almost immediately in the video, though I would've preffered something like "Ok ok, it's not literally every cable, there's no 6A, but I have 5, 5e, 6, 7 and 8" right at the beggining, but it's ok as it is too) but also that the one missing is rather important nowadays. Now, I know I said and asked a lot. Don't feel the need to mind me, I just provided the full feedback on my view, hoping it will be helpful. Other than that, I'm now very curious if Cat6 can do 10gbps at 100m. Do you have any plan to test interference too ?

[@thomasmaughan4798]

"literally the exact one that everyone should buy now." It is rare that anyone needs anything more expensive than CAT 5e.

[@Steamrick]

Without having seen the video, looking at the lengths - I bet that every cable hits 10Gbps. Cat5e probably manages 10Gbps at 'household' lengths such as up to 30m unless you've got something causing interference. There's a reason cables are typically rated for their speed at 100m length. I'll probably edit this after watching the video. edit: Yeah, as expected. I'd love a followup with 30m and 100m cables for cat 5, 5E and 6. No need to bother with the higher rated cables, they'll definitely manage.

[@landpet]

I have ordered cat5e and cat6 cables from short distance to 300ft (91m) so those videos will come as soon as I receive the items. The cat5 I didn’t see where to get the cables, so I’ll stick with the other two and maybe check a few more places for cat5

[@Steamrick]

@@landpet That's great to hear.

[@TrailBikeMike]

Super interesting. Loads of people here complaining the test is pointless because the cables are so short, but I disagree. I have a few longer lengths to APs and switches which are 6a. But I have loads of shorter lengths from switches and the router, to devices. A pity I saw this too late as I threw out all my 5 and 5e cables thinking they were useless. At least all my new cables are matching in white now.

[@darkdestroyer32]

I'm actually VERY surprised by this, but then again, it's likely based on the length being so short. It'd be interesting to see super long runs.

[@landpet]

I’m ordering the longer cables just to make that video. I’ll make it as soon as I get them

[@Kavyatej]

@@landpet thanks!

[@johnlehew8192]

Are the CAT 5,5e,6,7,and 8 cables all 3 feet in length?

[@hescominsoon]

you won't see the perfect 10 as you loose some due to protocol overhead.

[@landpet]

Makes sense

[@Aranimda]

Meanwhile, I have 150mbit/s down and 50mbit/s up for my internet and a gigabit switch using cat5e cables for my LAN. It's more than I need.

[@bren.r]

What monitor is that?

[@landpet]

It’s an LG OLED 48” C1. The new version is the C4

[@bren.r]

@@landpet no wonder it gets so bright - desktop OLED monitors don't get remotely that bright lol.

[@gathonar]

For anything under 15 meters cat5 will do fine with no degradation same as any cheap HDMI cable under 15 meters.

[@ArcadiyIvanov]

Importantly, you should be running them full-duplex for the speed test as you need to test for interference.

[@landpet]

Hmmm, unless there is an app that does that. I guess I can kick off the speed test on both computers at the same time. That would be a full duplex like.

[@bt81-]

I tested "every" ethernet cable... but I don't have a cat 6a.

[@nikolaialeksiev2536]

Different categories are pretty much advanced wire materials, insulation and coating that reduce signal loss and improoved connectors.

[@landpet]

I ordered various length cables up to 300ft (91m) for cat5e and cat6. New videos coming soon.