The trouble with the Samsung Electronics announcement last week of a 60 GHz version of a WiFi chip – usually referred to as WiGig – is that there are no details of what it will be used for and what type of device it can be houses in.
We have followed 60 GHz signalling for about 8 years, and people always talk about line of sight communication inside a single room, because 60 GHz is prone to interference through any single obstacle. But because there is so much bandwidth in this unlicensed slice of spectrum, if anyone can get it to work then it can potentially carry something like 7 Gbps.
Samsung has actually cited a slightly slower number and talks about sending uncompressed video around the home at speeds just under 5 Gbps, and it attracted headlines which made it appear to be a breakthrough pioneered by Samsung on its own. This is not the case.
The truth is that there has been an awful lot of work by many companies to exploit 60 GHz indoors and the first company that comes to mind is SiBeam, which drove an invention these days referred to as Wireless HD. This company spent $millions researching this and once it became likely that 60 GHz transmission might work in the home, the WiGig Alliance emerged trying to claim the spectrum for itself.
SiBeam was sold to Silicon Image one of the IP holders behind HDMI, for $36.5 million, after using up $112.5 million in VC funding. Today that product is sold as a way of simplifying home theatre installations without the need to put the set top by the TV. It boasts low latency and support for uncompressed 1080p/60 video through the air.
So it’s a bit hard to see that anything which Samsung has done is any kind of breakthrough.
The other major that is deeply into 60 GHz is Qualcomm Atheros who recently bought Wilocity, which itself came out of the 60 GHz work within Intel and between the two, Atheros pushing 5 GHz WiFi and Wilocity pushing 60 GHz designs, they built a single chip that can speak both types of WiFi. But that came out in 2011 and we have yet to hear of a single design win for it. Perhaps the Samsung chip takes the technology forward so that it can actually be used, but we would need to be told why and how this worked.
Already there are rivals for uncompressed video over wireless, such as Amimon, an Israeli start-up which makes most of the chips that are built into UHD wireless cameras, using something called WHDI. This is an entirely different approach to the problem and works in WiFi 5.0 GHz spectrum and it can send multiple HD signals around a home using MIMO antennas in a 40 MHz wide OFDM carrier, similar to 802.11n, although specifically not using WiFi protocols. Instead it works by splitting uncompressed video into two streams, which have different Forward Error Corrections applied to them. It claims to be able to send the equivalent of 3 Gbps of normal video in something closer to a 250 Mbps stream because of the way it treats that video.
The Samsung press release talked about the use case being sending video from a handset to a 4k TV set and claimed that “A 1GB movie will take less than three seconds to transfer between devices, while uncompressed high-definition videos can easily be streamed from mobile devices to TVs in real-time without any delay.” We’re not sure that anyone is actually going to do that.
But anyway it is a nice use case, however 60 GHz was initially notorious for the high use of electrical power that it demands, so this could be one of those apps that simply empties your battery in record time. Already there are Wireless HD chips that are supposed to be low power enough to be put into handsets and these have existed for some time, and yet we don’t know of anyone that has gone ahead and done it, so why would they do it with a Samsung chip – unless of course Samsung decides to use the chip. Samsing has a history of buying the best chip for a particular job on the market and that often means that Samsung’s semiconductor arm is left out in the cold.
The release also talked about this chip being a breakthrough in the sense that previous 60 GHz chips were line of sight, suggesting this one uses MIMO so that it can re-establish a different path around any obstacles (like humans moving around the room). Our understanding is that all 60 GHz chips use MIMO in the same way, so no breakthrough there.
The release said that “By leveraging millimetre-wave circuit design and high performance modem technologies and by developing wide-coverage beam-forming antenna, Samsung was able to successfully achieve the highest quality, commercially viable 60GHz Wi-Fi technology.” It also says that it uses the world’s first micro beam-forming control technology that optimises communications in 3 milliseconds when the environment changes and that it expects commercialisation next year.
Claims that Samsung can apply this to a wide range of products, including audio visual and medical devices, as well as telecommunications equipment and that it will be integral in the Samsung Smart Home, fill us with little more than scepticism. The likely outcome like all 60 GHz chips is that there will be no buyers and where there are, they will be in very small quantities for some time to come.
And going back to that fast read write use case in transferring video, back in 2010 Sony launched a technology called TransferJet which worked in 4.5 GHz spectrum for very short range connections just 1.25 inches away, supporting data rates up to 560Mbps and it would take about 15 seconds to transfer that 1GB movie, rather than 3 seconds and this is already widely deployed in Sony devices, so we think that use case is covered.