What does the Internet look like?


I think you should draw the internet. Yes, the whole internet.

Let’s back up. When we were little, we all liked to draw what we saw: trees, houses, people, etc. Yes, our drawings were rough and oversimplified, but they showed a basic understanding of the world. Even as adults, we like to use simple representations (like stick figures) to depict complex things (like people). We agree that a stick figure isn’t a great representation of the human body, but — when we need a quick visual placeholder — it’s a good start.

Unfortunately, from chats with friends and family, I’ve learned that most of us can’t actually draw the internet. Not even a stick figure equivalent. Some of us have drawn symbols — webs or clusters of interconnected dots — but nothing close to the internet’s actual, physical reality. That scared me a little. We should all be able to draw the ‘net, however crudely, to show that we know how it works. And it’s essential, as an informed consumer, to know how it works… but don’t get me started.

You might be thinking, “But, the internet’s not physically concrete enough to draw!” As Andrew Blum writes in Tubes: A Journey to the Center of the Internet:

“Two billion people use the Internet, in some form, every day. Yet physically speaking, it is utterly disembodied, a featureless expanse: all ether, no net. […] When the Internet first took off, in the mid-1990s, we tended to think of it as a specific kind of place, like a village. But since then, those old geographic metaphors have fallen out of favor. We no longer visit ‘cyberspace’” […] Instead, we think of the Internet as a silky web in which every place is equally accessible to every other place. […] The preferred image of the Internet is instead a sort of nebulous electronic solar system, a cosmic ‘cloud.’”5

But, the truth is, the internet is physical. Or, at least, it’s made from physical objects. Without computers, switches, nodes, cables, servers, and data centers, it wouldn’t exist.

So, to help you understand and draw the physical internet, here’s my best stick figure representation, followed by some brief terminological explanations.

Servers are basically computers, owned by companies like Google and Facebook, that store data.

Cables are physical wires that transmit data.

Tier-1 telecomm companies are major corporations that pipe internet around the globe.

Internet Exchange Points (IXPs) help various internet providers connect to one another, increasing the speed and availability of the internet.

Cellular carriers provide telephone/internet access to cellular devices via mobile signals.

Internet Service Providers (ISPs) provide internet access to devices via WiFi.

Devices are what we use every day: smartphones, tablets, and laptops that connect to the internet.

If that’s enough information, just pause here. Take out a pen and paper, and try to actually draw the internet! It’s much easier than it looks.

Or, if you’re already familiar with the basics of internet infrastructure, we can move on to the next level. Let’s flesh out the stick figure, adding facial features, fingers, and toes. My drawing below is a bit gnarly (not unlike a sensory homunculus), but it’s one step closer to reality.

Here we see a server again. It’s filled with all kinds of data, including websites. Even though data can display on a device in your hand, it technically lives on servers owned by other companies.

Most companies, especially large ones, need a lot of servers to house their data. So they build data centers: large warehouses full of servers. Data centers are actually a sensitive subject; Google Maps won’t even show the location of their main center. As Andrew Blum writes, “It’s become a cliche that data centers adhere to the same rules as the secret cage matches in the movie Fight Club: ‘The first rule of data centers is don’t talk about data centers.’” […] why all the secrecy […]? A data center is a storehouse of information, the closest the Internet has to a physical vault.”6 Given the value of our data, you can imagine the value of any given data center.

Data centers are located all around the world and owned by thousands of different companies. This means that the majority of our data doesn’t physically live anywhere near us; it lives… somewhere else… somewhere hidden and ineffable. Marketers have cleverly decided to call this “the cloud.” Cloud storage has a lot of benefits, but its fluffy name obscures its true physical reality. “[T]here is nothing cloudlike about it.”7

Data travels out from data centers to other parts of the world via intricate physical networks of cables and routers. Cables can transmit data in the form of electricity across metal (in the case of copper cables) or as light across glass (in the case of fiber optic cables). Internet cables usually run across preexisting physical infrastructures, like railroads, telephone lines, and power grids, since that makes installation easier. (Fun fact, “Sprint” is actually an acronym for Southern Pacific Railroad Internal Networking and Telephony.) And different pieces of the cable (like the outer tube, the glass threads, or the pulses of light) are actually owned by separate companies. As a packets of data travels across cables, routers help manage and direct its path.

A packet of data will eventually reach a Tier-1 telecom company, a major internet provider. In the US, the Tier-1 companies are AT&T, CenturyLink, GTT, Verizon Enterprise Solutions, and Zayo Group.8 Unlike local Internet Service Providers (#9b), these companies are so large and resourceful that they don’t need to get internet from anyone else; they just “peer” (#8) with one other and sell access to smaller companies.

Over the years, Tier-1 companies have created major cable pathways around the world, often referred to as the internet backbone. This backbone stretches across countries and even continents, thanks to large fiber optic cables laid along the bottom of the ocean. Here are some beautiful interactive maps showing the internet backbone and submarine cables.

As a packet of data travels along the backbone of the internet, across thousands of cables and routes, it will reach an Internet Exchange Point (IXP) at a colocation center, which is where Tier-1 companies, cellular carriers, and local Internet Service Providers (ISPs) connect to one another. By connecting in strategic locations around the world, these companies can increase the speed and availability of internet for their customers.

The act of Tier-1 companies, cell carriers, and local ISPs connecting in this way is called “peering.” In my drawing, it looks like peering is only taking place between about ten entities in a few combinations; in reality, peering is a ridiculously tangled web of locations, people, devices, and handshakes. Even trying to actually map it out feels like this.

Eventually, after traveling across many miles of cables, a packet of data might reach a cellular carrier. Most of us are already familiar with the idea of cellular carriers, since we pay them monthly for our cell phones. Keep in mind: some carriers have the same names as Tier-1 telecom companies (like AT&T and Verizon), and some have unique names (like T-Mobile and Sprint). Lower tier providers will have to peer with larger providers to get internet access and, all in all, the overlap in company names and services is part of what makes internet infrastructure so damn confusing.

Thanks to more cables and routers, a cellular carrier will pass the packet of data to a cell tower with an antennae on top. This antennae will transmits the data to your device via mobile signal. (When we don’t have any “bars” on our phones, it’s because we’re not close to a cell tower or because the mobile signal is being blocked by something.)

Alternatively, the packet of data might have gone through your Internet Service Provider (ISP). Again, most of us are familiar with the idea of ISPs, since we already pay them monthly for WiFi. Most ISPs also offer “bundling,” which means they’ll sell you phone, TV, and internet all together, since those services will be transmitted across the across the same cables anyway. Once again, some ISPs have the same names as Tier-1 companies (like Verizon and AT&T), and some have unique names (like Optimum and Spectrum). Lower tier providers will have to peer with larger providers to get access to the internet, and the overlap in company names/services causes plenty of confusion for the average consumer.

Thanks to more cables and routers, the ISP passes the packet of data to a Wireless Access Point. At home, this is probably your modem/router, which also serves as the hotspot: the point of transmission for WiFi.

Finally, the packet of data will travel across WiFi to reach your device. It might load on your internet browser or in an apps, depending on how you’ve requested it in the first place. You could also download the data to your hard drive (essentially turning your device into a personal server). Steps #9-13 are all intimately familiar for us, since we use devices every day; so I’ve labeled this section the “familiar consumer experience.”

However, lest we forget, the data still has to travel somehow from your device into your eyes and brain. This happens thanks to a physical interaction between yourself and your device: the device has a user interface (UI) and you have a user experience (UX).

Lastly, you might also have physical interactions with gadgets/accessories, like wireless speakers. If your speakers are connected to your smartphone via bluetooth, for instance, then your headphones have theoretically become becomes a part of the internet’s physical reality as well. In fact, the growing ideas that we could connect everything — headphones, cars, microwaves — to the internet is often called the Internet of Things (IoT).

Of course, this second drawing is much harder that the first. And, like most drawings, it’s still oversimplified and subjective. But if you give it a shot, congratulations, you’re drawing the internet!

Header image by Creaticca Creative Agency from Noun Project.

Back to the blog

  1. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (2-6).
  2. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (238-239).
  3. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (230).
  4. Tier 1 network.” Wikipedia. Web. Accessed on 2/6/18.
  5. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (2-6).
  6. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (238-239).
  7. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (230).
  8. Tier 1 network.” Wikipedia. Web. Accessed on 2/6/18.
  9. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (2-6).
  10. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (238-239).
  11. Andrew Blum, Tubes: A Journey to the Center of the Internet. HarperCollins, 2012 (230).
  12. Tier 1 network.” Wikipedia. Web. Accessed on 2/6/18.