One of the questions we have been trying to ask ourselves recently is what happens to set tops if the logic for them goes to the cloud. A number of people claim things on both sides of the equation. On the one hand the idea of the cloud not being ready yet, to do everything that a set top can; the other being that using the cloud can save as much as 80% from the total cost of ownership of a new generation of set tops. Both are clearly wrong.
A white paper that can be downloaded from the ActiveVideo web site which claims the 80% plus TCO gains has an element of fakery about it because it assumes that you use the existing set top to do the very little work that a new thin client requires. It compares an upgrade using the Cloud with no new set top at all, to one that has a completely new, top of the range set top. The comparison iin unfair.
While we entirely ‘get’ the idea of the cloud replacing the set top because it offers a refreshed UI, with new capabilities being added over time, rather than in one go when the set top comes out. We also understand that some work has to occur down where the set top is, in order to process the instructions coming from the cloud. That means a current generation set top might be left alone for one or two generations, but at some stage it has to be replaced with something, and no matter how cheap that might be, and such a replacement was never included in the TCO calculation carried out by ACG research, which provided the White Paper with ActiveVideo.
So we asked Murali Nemani, the new chief marketing officer at ActiveVideo, there since May this year and he talked us through his excitement at ActiveVideo, ‘The eco-system for cloud rendered UIs is starting to align, the market forces, the model, the readiness of the technology at both the server and the network, are aligning pretty well.’
Nemani, points out that what ActiveVideo does is all about delivering the User Interface as video and that at first this is over a managed network not the unmanaged internet, which he says only comes later.
Nemani says that ActiveVideo is the culmination of almost 15 years of work, but it came screaming onto the scene over the first installation at Cablevision, which was mostly to do with using the technology to render Apps in the cloud and offer 40 Apps on existing equipment. It was then that Verizon brought a lawsuit against Cablevision and ActiveVicdeo successfully countersued, landing $260 million in a US court room. And until last September that was more or less the news.
But if anyone thought that ActiveVideo was just a flash in the pan technology and that this was its zenith, they have another think coming. It emerged last Fall at Ziggo, the largest cable operator in the Netherlands, but once again that installation was never properly explained until it went live some time later. ‘Like I said this is about putting the video over managed networks first and then later unmanaged. Ziggo has used our system to reach its existing installed set top base. In the first 6 weeks of the system going live there was a 30% jump in the VoD volume of transactions, and 79% of that footprint were legacy boxes. The Ziggo CEO has had to make a statement about it in the first quarter’s accounts,’ said Nemani.
What the Ziggo CEO actually said was that in March it had launched the service on old Humax and Samsung digital receivers and that it has plans to extend this to include all its set top in due course. These older receivers weren’t even able to offer interactive services before, including VoD. Since the launch of the new UI in March almost 60,000 decoders were activated using the cloud-based solution. Prior to this only about 350,000 newer devices could take VoD services.
‘During the first quarter we again experienced a strong increase in the
number of transactions of over 80% year on year, with the penetration of interactive receivers also growing rapidly, but still relatively small compared to the total number of customers,’ Ziggo said. That pretty much confirms what Nemani was saying, that Ziggo had been able to add a new UI to 60,000 homes so they could use VoD and this only takes the total number of interactive customers that can buy VoD services to 410,000 out of a total of 2.8 million pay TV homes. If that could put VoD usage up 80% what would upgrading all of them do? And to be very clear to set top makers, so far this is 60,000 set tops that will not be ordered any time soon, and potentially it takes 2 million or more off the market.
‘Ziggo hasn’t even done multiscreen TV Everywhere yet with the system, although it has plans to,’ Nemani added. He went on to say that effectively this is a way to implement network PVR, and that across Europe, with the exception of Germany which has a rights law saying that nPVR is not allowed, this is now the company’s priority, ‘For us nPVR is an automatic catalyst to the move to the cloud.’
Then he wants to talk about Charter. ‘The first crack in the dam in the US was what we did at Charter,’ Nemani added, referring to the first US cable operator that has declared it will go down the Ziggo route, rather than what we might call the Liberty Global route. Liberty Global in the Netherlands got itself into a $100 million plus development (some say a lot more) of its Horizon set top, partnering Samsung, Intel and NDS and others. Comcast seems to have taken the same approach for its X1 set top.
‘People are starting to realize that we can change the UI for new devices like the XI using ActiveVideo from the cloud, all we have to do is port our secret sauce, which is a 200kilobyte footprint program. All it does is capture key strokes, and hand off to any onboard video decoder, and every set top has to have a video decoder built in. The thin client requires no native app development, if you want to put Apps on your TV either.’
In June ActiveVideo signed up Charter in the US as well as Sumitomo, a Japanese conglomerate best known for its banking successes internationally, to push the product in Japan.
Sumitomo has a holding in JCOM, the Japanese MSO that was once owned by Liberty Global, with control lying with KDDI, which also owns Cablenet. So KDDI controls around 8.3 million cable homes and most certainly could embrace ActiveVideo to bring them a new interactive UI. That might put another 8 million new homes out of the set tops market overnight.
ActiveVideo most recently introduced what is effectively an HTML 5.0 browser on any device, even if it is underpowered, by putting much of the processing into the cloud. HTML 5.0 needs a more capable Javascript engine, an upgraded GPU and 3D acceleration or it needs help from the cloud with the results delivered as video.
If an MSO like Charter can add apps and go to HTML 5 with a new UI, without changing a single piece of hardware, then at Faultline we concluded last month that there is going to be a massive hole in the order books of set top companies over the next couple of years. We think that other new capabilities, especially screen resolution upgrades, will mean that set tops will still continue to be changed, but on a less frequent basis, at a lower price and for different reasons.
‘The days of $300 set top economics are going to meet very challenging conditions,’ predicted Nemani, and he openly confirmed that there are already talks with Liberty Global on the technology despite its investment in Horizon and further he claims to be talking to a further 5 tier 1 operators and promises more major announcements by September. For its part Liberty Global has made it very public that it must consider a cloud solution next.
‘What we are proposing to Liberty Global is that it is not difficult to upgrade a lot of their legacy devices to the same resolution HD quality with the same UI, the same content and also to extend that to unmanaged devices outside the home. That doesn’t mean we are making our core proposition a version of this over unmanaged networks, the first instance is always over a managed network.’
Nemani makes no secret of the fact that Cisco has licensed its technology and that NDS, which Cisco bought, designed the Horizon device. Cisco seems to be planning to put this technology to work in its Telus win in Canada, hints Nemani, ‘That’s happening, as Telus moves from QAM to DOCSIS, the UI and the devices will separate. At LGI the browser is locked in because it has chosen Adobe Flash, but we can decouple that relationship.’
‘There are two ways to deal with the cloud UI notion, the conventional way is to have a device and use the cloud to send metadata, you have a browser in every device, and it gets tons of data, and you have to have enough processing power to process it all. Our alternative model is to have a very dumb thin client, the decoding, the browser and the rendering of the video are all done in the cloud. The entire thing is a video stream, and that can still be encrypted as it leaves the cloud server, in the same way as video always has been with conditional access. It is a simple API call and hand it off to the conditional access system before it leaves, the same way video always did.’
The implications for the market are that TVs themselves along with VoD servers and set tops, will all need to have a fraction of the time and effort put into then in future compared to today, if the cloud model is taken up. And ActiveVideo isn’t by any means that only company offering this type of cloud solution, although most others do it in a very different way, so the effect on set tops is not down to how well ActiveVideo as a company does, but how the entire genre does.
‘Since Charter everyone has been calling us. If we are not successful then everyone will need new set tops at the current rate, but if we are, then almost no new set tops will be needed,’ he ends.
At Faultline we take a more rounded view. People need gateways, and people need the processing power of set tops that are there, and there are other pressures for the introduction for more, rather than less set tops.
For instance the whole-home DVR idea where you have one intelligent gateway and lots of smaller, cheaper set tops around the home is leading to a proliferation of up to 5 or 6 devices in the home, as well as one super-gateway, with three times the processing power it ever had before. These types of devices also take on transcoding to export to portable devices on unmanaged parts of the network ‘ multiscreen.
Is the requirement for these going to go away? No but the amount of power needed in the gateway, if you embrace cloud, reduces once again. It was put there to manage transcoding and to run apps. If the apps run in the cloud and you simply stream video to the particular IP address represented by the tablet or smartphone or PC, and it has a very thin client on it, and manage the transcoding in the cloud too, you have no processing requirement to upgrade even the most basic of set tops. So those gateway designs, which also tend to manage WiFi streams etc’ will all need re-designing to a point where they reduce the bill of materials considerably.
And if there is a proliferation of tablets how long will there be any growth in zapper boxes for? One year perhaps? And this is accelerated once operators negotiate out-of-home video rights, which tend to run about a year behind the in-home rights discussion.
There are other reasons for new set tops ‘ things like new UIs, new video resolutions, new advertising insertion paradigms, the fact that they break, because DTT devices want storage and conditional access added, but for each reason, the cloud offers a counter balance, and these will more or less all be delivered sometime in the next two years.
So we can see that there might be a bumper year now, an okay year next year, and then set tops might slow. How slow will they go?
Well let’s simply assume that today the average life of a set top is about 4 years. If we skip simply one generation, and then the industry innovates to find reasons to replace devices even if they still attach to cloud and are very different, then we will have the same number of replacement shipments, but with a missed generation so they will be over an 8 year cycle instead of a 4 year one. That’s half as many, slightly simpler devices, in twice the time. Ouch.In the meantime items like this make planning at a set top manufacturer tougher.
Just to be slightly contradictory, the Infonetics quarterly set top service came out this week, and it says that set tops could not be healthier and it makes Cisco the new global leader.
Not only is Cisco now clearly in the lead in the market, having deposed both Pace and Motorola, leading as it has over each of the last four quarters, but also that the market has surprisingly not reduced in Q1 2013, despite seasonal weakness being usual at this time of year and the general concerns over set tops being on their way out.
While Cisco is clearly growing, especially within IPTV, others suppliers are falling away in Q1, in some cases by as much as 20% in terms of shipments. So if the market has fallen 5% since last quarter and cable not at all, you would have expected that fall to have come out of DTH suppliers such as Technicolor, which seems to have now fallen out of the rankings. IPTV continues to climb and Cisco now seems to be number 1 in both IPTV and Cable.
We suspect this is not so much seasonal, as cable is going through these development curves we have talked about earlier. For DTH suppliers this will be dramatic then. Given that India and Latam makes up huge growth markets mainly driven by DTH, and yet DTH set top volumes and revenues globally have both gone down, this must be partly due to falling ASPs due to increased business in these BRIC regions, but also we suspect this is a similar ‘design’ or ‘thinking’ pause, which cable went through a year or so ago.
DTH services like DirecTV have invested heavily in multi-room DVR, but now it probably believes that it needs unmanaged multiscreen from the cloud. It has to come out of one expensive device type and invest in another approach which targets tablets. And given that almost no DTH suppliers have their own managed networks, they have the prospect of delivering TV next over unmanaged networks, something that even ActiveVideo is not pushing that hard.
Even Infonetics admits that some of its numbers come about due to what’s going on in China and India, indicated by the fact that 56% of set top shipments were SD rather than HD.
Infonetics says that the global set-top box market in Q1 2013 taking in IP, cable, satellite and DTT STBs, and bolstered by adding OTT media servers, was up at $4.4 billion, down 5% from what is traditionally the strongest quarter in Q4.
Only cable set tops and OTT media servers actually gained on a sequential basis in 1Q13. It also points out that China Telecom just released a new tender for smart IP set tops, so once again China will come to the rescue of set tops, which apart from that, might become a moribund subsector for some time to come.
Infonetics claims that EchoStar has somehow pulled ahead of Technicolor in the set top market, but we think they are neck and neck given that the quarterly set top revenue of EchoStar was only $427 million last quarter. We figure they both have a 10% share, more or less.
But by including the money spent on OTT equipment, Infonetics appears to us to be including the type of upfront money that companies like ActiveVideo need in server capacity, and also part of the encoding and transcoding central head end revenues that in the past it did not include.
It seems to want to add what is probably as much as $500 million a quarter from the OTT market, and used the figures to artificially inflate set top revenues. If you take that off then the set top market looks like it is already in savage decline instead of doing $4.4 billion, doing something closer to $3.9 billion, with something closer to a 16% fall in the market, and that would give us a clear feel that the effects of the encroaching cloud are already being felt.