New CDN Vendors Still Launching: Updated List Of Content Delivery Networks

Over the past few quarters, more content delivery networks have launched services in the market or have announced upcoming offerings. To try and keep track of all the different companies, twice a year I update my list of content delivery networks who offer some form of video delivery services. (To make the list easier to find on my blog, all you have to do is go to www.cdnlist.com for the latest update)

Every time I publish the list I get many comments from those who work at some of the CDNs saying it is not fair to compare vendors on this list with one another. For the record, this list is not comparing one vendor to another. It is simply a list of the vendors who in one shape or another offer some kind of video delivery offering. It would be like making a list of all the car manufacturers and the list would contain both Ford and Ferrari. But that does not mean you are comparing a Ford to a Ferrari. It's just a vendor list.

When it comes to the term "CDN", it is one of those terms that is very generic and can be debated all day long, similar to others terms in our industry like "cloud computing" or "edge network". Everyone seems to have different opinions on what it should mean. That being said, take a look at what companies analysts and the media write about in the market and see who they do and do not reference as a CDN. Some may not agree with them, but that's reality. The market defines who they think a CDN is.

To use this list to make comparisons of one provider over another without looking at a company's size, products offered, revenue generated, geographic reach of network, number of formats supported etc.. would simply be inaccurate. Some of those companies listed are in beta with their offering and some literally have no customers as they have just launched in the market. Again, it is just a list of vendors, some of which are pure-play CDNs, others which aren't.

The companies listed are those who provide outsourced CDN hosting services. That's why companies like Konitki and others are not listed since they provide software to deliver video across your enterprise network. Same goes for companies like Swarmcast which is many times classified as a CDN when they are really a video platform. Additionally, companies like Ascent Media are not on the list as they provide video delivery service across a closed network from one customer location to another. For this post, I also did not list the video delivery providers who target a very specific geographic region which I classify as "regional service providers". At last count, there are over two hundred of these providers and you can find many of them listed in the StreamingMedia.com industry directory.

In alphabetical order these are the video delivery networks that I am currently tracking in the industry:

• Accelia
• Abacast
• Advection.NET
• Akamai
• Aflexi
  
• Amazon Web Services
• AT&T
• BitGravity
• BitTorrent
• Broadmedia
• CacheFly
• ChinaCache
• CDNetworks
• Cotendo
• Conviva
• Deutsche Telekom
• Digital Fountain
• EdgeCast
• EdgeStream
• Grid Networks
• Highwinds
• Internap
• Itiva
• Level 3
• Limelight Networks
• Mirror Image
• Move Networks
• NaviSite
• NTT Communications
• Pando Networks
• Panther Express
• PCCW
• PEER1
• Real Broadcast Network (RBN)
• Reliance Globalcom
• Tata Communications
• Technicolor
• Velocix
• Voxel.net
• Vusion

If you think a company is missing from this list, please feel free to add it in the comments section and please make sure you read all of my post before making any comments.

Comments

[grinsandfun]

why don't you make the list the true players vs. resellers? Example - Tata = BitGravity, Move Networks - more of a player technology and reseller of L3/AKAM/LLNW - not really a CDN per say..

[Dan Rayburn]

Tata is not considered a re-seller since they license the BitGravity technology and deploy it on their own network.

As for Move Networks, you are right there. I always debate whether or not to put Move on the list as they really are not a CDN - today.

[John]

www.qbrick.com

[Nancy]

Check StreamZilla.eu
- Triple A+ global performance
- Global customers
- rolling out servers outside europe

[Dan Rayburn]

Both Qbrick and Streamzilla have no footprint at all in the U.S. and while many networks call themselves global, you can't be global if you don't have any actual servers in North America.

[Elgar]

Brightcove....

[Dan Rayburn]

Brightcove is not a content delivery network and has none of their own infrastructure for delivering video. Brightcove sends their customers content over Limelight and other CDNs.

[Laurent]

Ipercast International

[Dan Rayburn]

Ipercast has no delivery in Asia and only one footprint in North America, located in NYC. I give them credit for not calling themselves "global" anywhere on their website.

[Nancy]

So if a CDN has global customers, competes with global CDN's but has no physical servers in the USA, even though the performance is as good, it is not a CDN?

Maybe that rule could be used five years ago. Because the networks were slow, you needed servers everywhere.

But today global networks are becoming so fast and large that the need for servers on each continent is becoming less important. The true bottlenecks are the local exchanges and the metro rings. It does not really matter anymore whether a CDN has a big pipe or a big pop.

It is a matter of volume and costs. Operating a remote edge pop can be expensive, especially in Asia: operating costs are high and traffic costs are extremely high. It is better to have a big pipe and have the servers in Europe, for example. Same performance, lower costs. Still a good CDN service.

[Dan Rayburn]

If a CDN does not have servers for video delivery in multiple continents, then they can't be global. The idea that you can still be global and have the "Same performance, lower costs" is something many want to say, but have no proof of.

How you can say "It does not really matter anymore whether a CDN has a big pipe or a big pop" makes no sense. If a customer needs to support 100,000 users, and the CDNs pipe can only support 50,000 users, how can you say the size of their capacity does not matter?

Many CDNs all use tons of marketing words to say they can do one thing or another, yet don't back it up with real facts to prove otherwise.

[Matt Levine]

Nancy,

Unfortunately the size of the pipe doesn't dictate tcp performance (streaming udp is fine, however) - TCP is a sliding window protocol, and performance is a function of pipe*latency. Most OS's still ship with a TCP window max of 32kB - it doesn't matter how big your pipe is if you're 100ms away.

[Nancy]

When you have a pop in Europe that can handle 100K viewers and a pipe that can handle 100K viewers, and the latency can be kept low, then there is no physical regional edge pop required.

By the way, the number of clients that actually need these volumes is small. Not every CDN customer is Disney or HBO. Think about the hundreds thousands enterprises and small businesses out there that need a good CDN with great global performance and great features but do not need that concurrency.

StreamZilla backs up its claims: Keynote did independent performance testing from 24 sites around the world and gave a triple A+ status for global performance: for every streaming protocol: http://www.streamzilla.eu/performance.html

These performance tests are done on the busiest lowest service level service within the CDN, without any optimization. All streamed from Europe. Only China underperformed, due to the government firewall.

[Dan Rayburn]

Keynote has given every single CDN that is in the certified Flash Video Service Provider program an A+ rating. I don't think that anyone would debate that Streamzilla does not have the reach or capacity that an Akamai or Limelight has. But since that is the case, how can they both get the same network rating? Does not make any sense.

One one hand the regional service providers want to showcase how they are different from the larger CDNs and go after smaller sized customers, are more focused and more flexible with things like commits and contract terms. But on the other hand they all want to be compared to the big CDNs at the same time. Well, which one is it? They can't have it both ways.

[Justin Foster]

I think the reality of the CDN market is that it's almost completely commoditized. Sure, you've got players like Akamai able to charge more and not every CDN is global which enables some ability to differentiate on something other than price, but to me and lots of business folks (who will play an increasingly important role in the selection of CDN partners moving forward) choosing a CDN is mostly a 'who cares' kind of decision. Content delivery is only one component of a larger video infrastructure. In fact, many apps will let users plug and play with multiple CDNs and some already do. The future of video is about applications, not infrastructure. That's the paradigm shift that needs to happen in this industry. A bunch of IT guys making decisions about this CDN or that CDN isn't what the future will look like; it'll be dictated more by business users seeking out specific video application functionality. As long as a CDN meets a basic list of criteria and IT has a seat at the table in the decision-making process and is able to check off some basic CDN criteria, that'll be the extent of IT involvement in lots of CDN decisions. Just my 2 cents. Getting off soapbox now :-)

[Nancy]

Dan, it's not that black or white with global or regional CDN's. I don't claim that the smaller CDN's have the same infrastructure. There is a big market that does not need hundreds of gigabits, but they do need a professional CDN with global reach and good performance. Too bad no other CDN actually publishes their Keynote results, are they afraid of being compared?

Justin, that is exactly the issue: today CDNs only compete on capacity and price. Most smaller CDNs try to compete the big ones by even lower rates and cut costs by offering less service. Look at their websites: most of them claim that their key advantage is a low price. If that is their main USP, they can prepare for a rough ride, especially if they over invested in hardware.

Smaller CDNs can offer similar performance but not the capacity. The true distinction is in flexibility, features and support. Compare it to cars:

When you want a big car that drives you and your family from A to B, has four wheels, is affordable and is fuel efficient, you get a Japanese car.
When you want a small car that drives you and your family from A to B, is dirt cheap but doesn't come with all the security features, you get a Chinese car.
When you want a fast car filled with gadgets, superior build quality and great service, you get a German car.
Sorry, no US cars in my list :)

Or another analogy:

When you want a commodity computer, buy a Dell with Windows. It does the job. Large CDN's.
When you want the cheapest possible computer, buy PC parts and have fun(?) getting Linux on there. There is no guarantee but it should work. Price war CDN's.
When you want ease of use and a quality computer, you buy a Mac. Maybe a smaller market share but higher margins. Exclusive CDN's.

[Dan Rayburn]

Yes, it is black and white with CDNs. If you don't have servers located in North America for instance, you can't call yourself a "global" CDN and expect to be taken seriously. You are right, not everyone needs a CDN that has this, but then why do you want to be compared to a global CDN when you aren't and are offering a different level of service? You can have it both ways.

Many other CDNs have put out releases about their Keynote results including Akamai, Limelight, Mirror Image, UpStream Networks etc.... not to mention many of them list those results in their SLAs. So the idea that Streamzilla is the only CDN to do this is false. A simple Google search shows you otherwise.

In one post you said smaller less regional CDNs could offer the same performance as other CDNs, but now you say "similar performance". Similar is not the same.

You also say that "today CDNs only compete on capacity and price". That could not be further from the truth. Some CDNs compete on that, but not all. Many don't.

[norbert]

hi there,

we would love to be part of your list. we are the german pioneers for streaming since 1997 and run our own cdn. at the moment we are the largest independent cdn in germany and the first one worldwide as far as we know that runs it's servers with energy taken from renewable sources.

we call it green power cdn

regards,

norbert

[Edward]

Beijing Fastweb Technology (www.fastweb.com.cn) is a major CDN provider in China, over 200 customers including Sina, PPlive, Ku6, Kingsoft, Reuters, Financial Times, Phoenix TV, 1-ting, etc. ....

[Ken Vodicka]

Dan, as you may know the A+ is derived from Keynote’s Streaming Perspective StreamQ™. StreamQ™ is a metric that has been around for years and is a standard for many companies measuring the streaming media user experience. This metric is also being widely used as a barometer for SLA’s. Please note that not all customers are receiving A+ letter grades today. Many of the certification customers started off with lower grades but made the necessary availability and performance changes to achieve an A+.

StreamQ™ is measured in letter grades from A+ down to F. This grade is based on the amount of time a user had to wait or experience interruptions versus the time actually spent enjoying the content. StreamQ™ is calculated based on the sum of the Connect, Buffer and Rebuffer times as well as the number of Rebuffer events that occur during the check.
Frustration Time = Connect Time + Buffer Time + ReBuffer Time.
Frustration time is made up of these three non-viewing activities. The number of Rebuffer events is also recorded and a one second penalty is added to the Frustration time for each rebuffer event. In simple terms, the total Frustration time must be less than 6 seconds to achieve an A+.

For additional information on Keynote’s Streaming Perspective please see http://www.keynote.com/products/voip_and_streaming/streaming_performance/streaming_perspective.html

[celine]

I am looking for regional CDN in North America. Where can I get a list?
thanks
Celine

[Taeha Park]

Please add KT to the CDN list. KT is the leading CDN provider in Korea since 2006.

Korea Telecom (partnered with Solution Box).

http://www.ktics.com (KT CDN web site)
http://www.solbox.com (Solution Box Inc.)

Thanks.

taeha