It can be surprising to see a giant like Intel announcing in San Francisco, where the WebRTC actors are plentiful, an IoT + WebRTC collaboration with a Korean Telco operator (zdnet, koreatime) . I think it makes sense from at least two angles: the x86 vs ARM, and the Asian angle.
Intel has always been perceived as a desktop CPU maker. Not GPU. Not embedded hardware, leaving smaller appliances to ARM. Their IoT module, the Edison, is quite big compared to other solutions. In an interesting reversing of power, Intel announced that he has licensed ARM IP to be able to provide ARM chips made using their 10nm foundries.
While Intel has a very strong and capable webRTC team, they are developing free (but not open source) software optimized for intel hardware. Their clients are companies that want to develop a webrtc solution or PaaS without having to code the entire infrastructure. More in a following post. Their goal is obvious: selling Intel hardware, but it also means they do not have their own webRTC PaaS, and they also do not have their own cloud offering.
It is hard to be a Telco in Asia (outside of china mainland which is a kind of special case): you are living in the shadow of the Japanese Giants. NTT is #4 in the world by total revenue (just shy of 100B USD), Softbank is #6 (75B), KDDI #13 (37B). The first Korean, KT corp, is #17 with 20B, and SK Telecom is #21 with 15B.
One way to make a difference is to innovate, and SK telecommunication has been pushing a lot on new cloud technologies, IoT and webRTC. For those of you who follow me on linked in as well, you might remember that my august 2015 “webrtc ecosystem” review was already mentioning them, and the attraction webrtc players had to IoT. However, it is still hard to compete with NTT, whose webRTC platform Skyway was released more than a year before that, and who is also a cloud operator. Everybody is also providing their solution for free, in an effort to attract developers.
On july, Softbank announced it intention to buy ARM holdings. It brought the technological core of IoT and mobile hardware into a company that has the mean to grow it fast and aggressively, as well as to develop technologies based on it, potentially a problem for Intel. That also reinforces the Japanese ecosystem, potentially quite a problem for SKT. Intel and SKT had a common problem, and matching technologies, always a good start for a collaboration.
Interestingly both IoT and webrtc share a lack of specification for the communication / signaling layer, which means that they in turn share a common need for one. While IBM’s MQTT seems to have achieved an industry-wide de-facto standard status for IoT signaling, as illustrated by AWS out of the box adoption of that standard, if you do have an existing webRTC PaaS, you can reuse the signaling part of the infrastructure for your IoT solutions. Fair enough. Intel who did not have a webRTC PaaS get it from the collaboration.
SKT playRTC, a little bit like NTT skyway, is heavy on the client side of things, and slightly lacking on the infrastructure side. They actually followed the same, usual, development path: first a web app/sdk and a signaling server, then a mobile app/sdk. Intel could contribute extremely good infrastructure pieces, with hybrid media server (SFU/MCU), SIP gateways, etc …. Plus they have some of the best T-Shirts in the ecosystem. (picture to come in a following post)
Interestingly, the INTEL webRTC team is based in Najing, China, making this a beautiful Asian fairy tale, and possibly a success story.
This winter (tentatively 28, 29th october, in Beijing), the Real Time Web Solutions conference will happen for the second time in China. It had over 600 attendees last year. Known speakers for 2016 include Baidu, Alibaba, and MoMo. Let’s hope that SK Telecom and INTEL (who had a one-day event co-located last year) will also be there and show us more about IoT and webRTC. Note for the non Chinese speakers: real-Time web Solutions conference provide both English and Chinese tracks, with translation. Intel 2015 one day event was Chinese only.
This work by Dr. Alexandre Gouaillard is licensed under a Creative Commons Attribution 4.0 International License.
This blog is not about any commercial product or company, even if some might be mentioned or be the object of a post in the context of their usage of the technology. Most of the opinions expressed here are those of the author, and not of any corporate or organizational affiliation.