Tools for testing your Internet speed

Topic

New Reply

	TOP STORY Tools for testing your Internet speed
	By Patrick Marshall There are many reasons an Internet connection will become noticeably slower — including not getting the bandwidth you’re paying for. Internet speed-testing services might help reveal whether your ISP is at fault, but only if you understand how they work.

---

The full text of this column is posted at windowssecrets.com/top-story/tools-for-testing-your-internet-speed (opens in a new window/tab).

Columnists typically cannot reply to comments here, but do incorporate the best tips into future columns.[/td]

[/tr][/tbl]

Reply | Quote

Replies

Understandably the test sites are US-Centric.

For UK members I’d recommend http://www.thinkbroadband.com/

Reply | Quote

I sometimes use http://www.speedtest.net and for those who want a more in depth look at their wireless connection which includes a speed check, you can use Netalyzr (requires Java – which has just been updated) http://netalyzr.icsi.berkeley.edu/ but speed tests are better done when wired.

Reply | Quote

It’s worth pointing out that no matter what you think you should be getting, the ISP’s don’t really lie about speed. The reason is in the phrase, “up to”.

ISP’s advertise speed as being “up to 15 mbps” or “up to 50 mbps”. That puts these numbers at the upper limit of what you can expect to get, not the normal or average. 5 mpbs is up to 50 mbps.

Reply | Quote

Good basic article. Although I was expected to see a reference to TraceRT (on Windows run from a command promt “Tracert URL”) Which will give you a good idea where bottlenecks are, once you determine your speed is slower than you contracted for.

Basically the options are: Between you and the ISP, the ISP, and beyond the ISP. Calling the ISP for something not owned by, or contracted by them is futile. However if TraceRT shows long ping times before your packets leave ISP servers, then calling the ISP is your only option. And having a trace in hand is very helpfull to the ISP in finding a starting place to look for issues on your line.

Good Luck getting though the ISPs CSRs to to the operations center to find someone that undertands what a TraceRT is.

My personal experience, I was using a DSL ISP that performed alright, latency crept back into the line usual reason being from them oversubscribing servers in the operations center. This time the trace showed a constant 50ms lag in the 2nd hop (my line to the operations center of the ISP). I called the ISP trouble line and was informed they would not refer the problem to operations until my latency hit 300ms. I left that ISP 5 days later.

Reply | Quote

I have a SamKnows Whitebox that constantly monitors my internet speed. The service is free including the hardware. I’d like to see this service addressed in comparison with the other speed tests noted.
https://www.samknows.com/

Reply | Quote

I have a SamKnows Whitebox that constantly monitors my internet speed. The service is free including the hardware. I’d like to see this service addressed in comparison with the other speed tests noted.
https://www.samknows.com/

You might want to read this.

http://www.farces.com/samknows-nothing-much-at-all/

Reply | Quote

I have found testmy.net to be one of the best speed tests. It gives minimum and maximum as well as the aggregate speeds. If you have the time you might also read their page “What makes TMN different?”

In addition, as a general rule I would make sure you are testing against at least a few locations that are not on your ISP’s network. The speed of your last mile connection and on the ISP’s network is certainly important but equally important is what kind of peering connections your ISP has with other ISPs and networks. I would even do a traceroute to see what the path is to the speed check server. Call me cynical but I would not put it past some ISPs to give preferential routing to speed check servers on their networks or to ones they recommend you to use.

Reply | Quote

What’s wrong (if anything) with using “Ethernet Throughput” under the Performance Tab of Task Manager in Windows?

My Rig: AMD Ryzen 9 5900X 12-Core CPU; ASUS Cross Hair VIII Formula Mobo; Win 11 Pro (64 bit)-(UEFI-booted); 32GB RAM; 2TB Corsair Force Series MP600 Pro 2TB PCIe Gen 4.0 M.2 NVMe SSD. 1TB SAMSUNG 960 EVO M.2 NVME SSD; MSI GeForce RTX 3090 VENTUS 3X 24G OC; Microsoft 365 Home; Condusiv SSDKeeper Professional; Acronis Cyberprotect, VMWare Workstation Pro V17.5. HP 1TB USB SSD External Backup Drive). Dell G-Sync G3223Q 144Hz Monitor.

Reply | Quote

Indeed. Here in the UK a certain percentage of the ISP’s customers must be able to get the advertised speed.

My ISP (Sky) advertises up to 17Mbps for their ADSL2+ product but under ideal conditions speeds of 24mbps can be obtained. Coincidentally I actually get their advertised speed.

Another thing to take into consideration is throughput vs synch speed. A rule of thumb for ADSL2+ is that throughput should be around 84% of the synch speed. In my case that means expected throughput of ~14.3Mbps. I actually get 14.5Mbps.

Reply | Quote

I used the tools mentioned in the Test Your Speed article on my gigabit fiber optic home connection provided by Vermont Telephone (VTel). I got wildly disparate results for a wired connection to my modem:

GeekSquad: 20.28 down 10.48 up
Ookla: 750.98 down 684.46 up
SpeedOfMe 99.35 down 33.67 up
Visualware 264 down 171 up
BandwidthPlace 73.44 down 33.63 up
CNet 102.85
Xfinity 588.74 down 649.07 up

Looks like they’re not set up to handle gigabit connections.

Reply | Quote

My Comcast here in NH advertises 105 Mbps. Using either Speedtest, Speakeasy, or Comcast testers gives me about 90 to 94 Mbps. But if I bypass my router and plug the ethernet cable directly into my computer I will get 105 or very close to it- to my surprise. Just checking; I’m very consistant with the 3 testers I mentioned – Download – 94.75, 94.91, 93.14. Upload – 12.10, 12.04, 12.08. (all via my router) I’m either lucky or just getting what I pay for.

Reply | Quote

My Comcast here in NH advertises 105 Mbps. Using either Speedtest, Speakeasy, or Comcast testers gives me about 90 to 94 Mbps. But if I bypass my router and plug the ethernet cable directly into my computer I will get 105 or very close to it- to my surprise. Just checking; I’m very consistant with the 3 testers I mentioned – Download – 94.75, 94.91, 93.14. Upload – 12.10, 12.04, 12.08. (all via my router) I’m either lucky or just getting what I pay for.

Depending on the max speed of your router (IE: 100 Vs 1000) that could be your very top end limiter for only getting 94 Vs 105.

On my wired machines I can speedtest or Xfinity test very close to the max 105 or better (and am going through router) when the network is not loaded, but on my laptop via wireless I won’t hit over 40, even though the laptop & router rated wireless speeds are much higher & are only about 5 feet apart.

On my work machine on same router through corprate VPN, only hitting 40 – 45 on a good day so plenty of room left the the rest of the house :rolleyes:
(up is only 5, so when uploading a 1 gig file, time for lunch or dinner!)

Reply | Quote

Depending on the max speed of your router (IE: 100 Vs 1000) that could be your very top end limiter for only getting 94 Vs 105.

On my wired machines I can speedtest or Xfinity test very close to the max 105 or better (and am going through router) when the network is not loaded, but on my laptop via wireless I won’t hit over 40, even though the laptop & router rated wireless speeds are much higher & are only about 5 feet apart.

On my work machine on same router through corprate VPN, only hitting 40 – 45 on a good day so plenty of room left the the rest of the house :rolleyes:
(up is only 5, so when uploading a 1 gig file, time for lunch or dinner!)

My router is a D-Link DIR-655 Gigabit, which has 1000 Mbps speed. I’ve read that any router will cost the user a little speed, and it seems true. Also, I just checked my wireless speed on an ASUS Zenbook- all three speed tests resulted in download 73 to 78 Mbps/upload 12. I’m not going to try to figure out why my tests are much more consistant than others, as the author of this article reports. I’m satisfied.

Reply | Quote

“In the future, ISPs might have to revisit download/upload bandwidth allocation.”

For cable providers (that is traditional cable) the upload channels were placed in a relatively small frequency band
so that they could use the same cable for both download of data and television and upload of internet data. The
distribution amplifiers were designed to follow suit. After you have reached the capacity of that upload band, there
is very little that you can do without redesigning the entire cable plant and replacing a lot of equipment at both
the customer premises and in the street. That costs big bucks.

FIOS and other optical providers may have provisioned their networks differently and have wider bandwidth to
play with.

Reply | Quote

I’ve used testmy.net for … 10 yearsish?? It doesn’t require anything but a naked browser (works even in old browsers, and without flash or javascript). It keeps track of your results for some time even if you don’t bother to make an account. (I assume it’s looking at MAC address since it has results for my same PC from three different cities.) Results are pretty consistent and in line with perceived speed (or lack of it).

Reply | Quote

Isn’t it important for us to test our contractors for internet speed, and our modems and routers for potential failures on a regular basis?

Is there a service I can subscribe to which will automatically check my internet connection and speed on an hourly basis, and then let me know when there is a problem?

Alternatively, is there a modem or a router which can check its eternals, or some other way to check these devices to make sure they are working properly instead of me arriving home from work and finding out my modem just croaked?

Many of us depend on these devices and services for our livelihoods. Currently, my only safeguards see to be always have a backup modem, router and alternative ISP on hand. I would like to know of tools/services which can warn me of trouble. Car’s have CHECK ENGINE LIGHT’s, TIRE PRESSURE LOW, and LOW OIL warnings. Can’t my “high tech” services/devices provide similar warnings…

Reply | Quote

Hard to know what to take seriously. With Opera 12.17 on an i7-2660K desktop computer, Charter uses OOKLA and returns 66+ mb/s down, 4.4 up and ping 9 ms, while OOKLA’s own site, depending on the server chosen, is running 4-8 mb/s down, about 2.7 up and 150+ ms ping. My experience is that file downloads often run at 7-8.6 mb/s.

Reply | Quote

Hard to know what to take seriously. With Opera 12.17 on an i7-2660K desktop computer, Charter uses OOKLA and returns 66+ mb/s down, 4.4 up and ping 9 ms, while OOKLA’s own site, depending on the server chosen, is running 4-8 mb/s down, about 2.7 up and 150+ ms ping. My experience is that file downloads often run at 7-8.6 mb/s.

If you are using Windows then see what the speed tests are like in IE, but you could also see what Netalyzr gives.

I’ve always found that gives pretty close to what my ISP speed tester gives.

Netalyzr requires Java and enabled in browsers.

Reply | Quote

Test with a hardwired connection if you can. My provider’s tech dude (who seems to really know his stuff) told me that using wireless generally cuts throughput by about 25%.

The problem with speed tests that require flash, java, or javascript, is that they’re somewhat dependent on your computer’s performance, and can be significantly laggy if the browser or system are ‘busy’. That’s another reason I like testmy.net — all it does is a pure data upload and download test. Your system’s performance is not a factor.

Reply | Quote

I’ve always assumed ISPs probably manipulate their QoS settings to prioritize connections to the well-known testing servers. Seems like when I run these tests, they often start accelerating as the test goes on and I realize there can be lots of reasons for that like warmup, but still…seems like a likely thing for the ISPs to be doing, doesn’t it?).

Reply | Quote

I suspect it’s more often that when you do run a test, that’s all you’re doing.

If that weren’t the case, CenturyLink oughta be embarrassed that I only get 15% of their national average.

Reply | Quote

I suspect it’s more often that when you do run a test, that’s all you’re doing.

If that weren’t the case, CenturyLink oughta be embarrassed that I only get 15% of their national average.

Have you ever thought of voting with your feet and look for another ISP ?

You don’t say what type of broadband you have – ADSL degrades with distance from the exchange and there’s nothing any ISP can about that.

Reply | Quote

Have you ever thought of voting with your feet and look for another ISP ?[/quote]

Great to have the choice! I can pay $15/mo. for 1.5Mbit DSL (it would be $40, but I raised hell about their lack of better options and got offered a discount), or $70/mo. (plus $200 installation fee) for 4Mbit fixed wireless, which works poorly or not at all in bad weather (this is Montana; the weather is often bad). Or I can pay $20/mo. for a dial tone and use dialup. Yep, great choices!

You don’t say what type of broadband you have – ADSL degrades with distance from the exchange and there’s nothing any ISP can about that.

ADSL. I’m about 16,000 feet out, but the problem is less distance than tired old equipment, and no competition worth noticing. CLink could upgrade the local exchange, but so far they’ve declined to do so. It serves ‘only’ about 10,000 people; apparently not worth doing.

Reply | Quote

If your computer is connected via WiFi to a home router, be sure to run a separate speed test while directly cabled to your router. It’s easy to overlook the speed degradation that might be occurring within your own home. Always a good practice before berating your ISP service rep.

Reply | Quote

That’s about 4.8km and would give you a Downstream Attenuation of ~67dB – no wonder your speed is crap.

Here are some others to look at http://montana.theispguide.com/ but your distance from the exchange is the killer and it’s the “up to” speeds that will be indeterminate until you are connected – not sure what Comcast are promising though ?

Reply | Quote

Yeah, I’m on the edge of DSL working at all.

Already called all those on the Guide page, plus all the local outfits I could find, no joy A couple (I think Comcast and Verizon) have their internet dept. phones auto-redirect to CenturyLink!! There’s no cable within a couple miles anyway. — DSL isn’t going to be any better over the phone wires til the exchange is upgraded, which might be never, and maybe not then. Hope to convince one of the less-expensive fixed wireless outfits that they should put a radio on my hill and serve all of us on this side of the highway who are currently blocked, but who knows if that will go anywhere.

Reply | Quote

Yeah, I’m on the edge of DSL working at all.

Hope to convince one of the less-expensive fixed wireless outfits that they should put a radio on my hill and serve all of us on this side of the highway who are currently blocked, but who knows if that will go anywhere.

Those ‘point-to-point’ wireless outfits are your best bet for getting decent broadband speeds in rural areas. CenturyLink, Comcast, or AT&T aren’t going to provide it for you, unless they’re forced by federal legislation.

In Finland, broadband access was made a legal right in 2009, even up in Lapland. Think about what that type of telecommunications reform could do for Montana.

Fast Internet access becomes a legal right in Finland
http://www.cnn.com/2009/TECH/10/15/finland.internet.rights/index.html?_s=PM:TECH

Reply | Quote

Satellite Internet looks like an expensive option and probably more just suited to businesses.

http://www.speedtest.net/local/montana

I wouldn’t like to think I was paying those sort of prices.

Reply | Quote

From that link (where their speedtest didn’t like my flash), “Satellite Internet, such as HughesNet and Exede, is a popular choice in Big Sky Country, as many residents are far from the beaten path.”

What a load. No one “chooses” satellite if they can avoid it, not at $80/mo. for 1Mbit, and a 10GB/mo. data cap (which was the best any of ’em quoted me). Rather, there are a lot of rural areas where it’s either satellite or dialup, with absolutely NO other options (and no cellphone coverage either). Little local fixed-wireless outfits are on the increase, but they still don’t cover everywhere. Folks who haven’t lived in the ‘flyover’ states generally have no idea just how BIG the “Big Sky Country” really is. A lot of the people who grow your food are 50 miles from the nearest town that’s more than a grain elevator and a rail siding. (Consider that I’m 15 miles from the largest city in Montana.)

There are still fixed wireless outfits charging $60/mo. for 0.7Mbit (I got quoted that by at least two of ’em, if they’d had coverage here at all!), which goes to show how that field needs a whole lot more competition.

I had fixed wireless in the SoCal desert (cuz my choices were that, or dialup, and if my house had been 50 feet further west I couldn’t have got it either) and my provider was a one-man band, and he liked to talk about his business. He purchased bulk bandwidth from AT&T at free for download, 5 cents/GB for upload. His Motorola radios were something like $700 for the transmitter and $100 for the receiver. I don’t know what the rest of cost-of-entry is but this is sure not out of reach. And there I was paying $39/mo. for 1.5Mbit (in good weather; it didn’t work at all when it rained). He also said the real objective was to get big enough to sell the business to one of the big names.

Reply | Quote

While we in the UK have similar problems in the rural areas and in some cases communities have clubbed together to install their own fibre cabling, I would have thought that in a country like the US and its technology, more would be done to help those in outlying areas – but with businesses – unfortunately it appears to be the same old story of the bottom line.

Reply | Quote

The thing is, Reziac does have broadband – just not very good.

Reply | Quote

That’s why a minimum Internet speed must specified. Finland started with a universal minimum of 1 Mbps in 2009. Now they are going for 100 mbps by 2015 – for ALL citizens.

Finland: Plan for universal 100Mbps service by 2015 on track
http://arstechnica.com/business/2012/10/finland-plan-for-universal-100mbps-service-by-2015-on-track/

Reply | Quote

That’s why a minimum Internet speed must specified. Finland started with a universal minimum of 1 Mbps in 2009. Now they are going for 100 mbps by 2015 – for ALL citizens.

Finland: Plan for universal 100Mbps service by 2015 on track
http://arstechnica.com/business/2012/10/finland-plan-for-universal-100mbps-service-by-2015-on-track/

That would be a pipe dream for us in the UK, although the Gov. is supposed to be helping out with getting high speed broadband to the far reaches – it just doesn’t seem to be doing it very well or fast enough for those affected.

@ Reziac – Suadi Arabia have towers 50 miles apart to provide 4G but they don’t seem to have any qualms about spending money.

Reply | Quote

In rural areas it’s not so much the almighty bottom line; it’s more the sheer distance involved, and that all physical cabling costs by the foot — not so much the wire itself but the labor to install it. (Last year I was quoted — are you sitting down? $75,000 per mile to bring in electric service.) Montana alone is three times the size of England, almost 700 miles from one corner to the other, and a great deal of it makes your roughest country look tame.

What baffles me is that the big providers don’t sidestep all the cost and bureaucratic hassle of laying fibre or copper, and go with big-capacity fixed wireless on their existing cell towers. That would bring better internet at least to the less-remote rural areas, at far less cost to everyone. How much of that is FCC regs impacting what they can and can’t do, I don’t know, but about 15 years ago Earthlink tried to bring fixed wireless into rural Los Angeles County (where I used to live) and that project got shot down.

Reply | Quote

Saudi Arabia is an anomalous economy. They have money to throw at all sorts of stuff the rest of us don’t. Shoulda invested more in those oil wells!!

Legislating broadband speeds is all dandy in theory but Finland is a little smaller than Montana, and is relatively flat which does wonders for both ease of construction and line-of-sight broadcasts. Where I lived last year, line of sight meant a few tens or hundreds of meters, cuz there were mountains in the way in all directions. (No cell service either.)

Reply | Quote