Snapdragon 180Real consumer speeds of hundreds of megabits will be practical for good connections beginning in 2017 some places.Update 1/3 I was wrong. Few reporters caught the gigabit. Qualcomm chose to emphasize performance (++30%?) and battery life. I assume they decided not to discuss gigabits until deployed more widely. end update "Gigabit" will expand rapidly to dozens of telcos around the world in the next two or three years, although the "gigabit" is a shared speed. Nobody believes it yet, but new technology can offer wireless abundance. Anywhere the market is working we're going to see wireless offers of five and ten times the current speeds and caps at 50-150 gigabytes. mmWave 5G is on top of that, one reason the telcos are unsure there will be enough of a 5G market where landlines are robust.

I thought Tuesday's Qualcomm presser would (finally) clue in the general press. Wireless experts have known this was coming since 2010-2012, when the necessary technologies became part of the LTE standards. Didn't happen, as Qualcomm didn't emphasize or explain how important this will be. It will take time. If the chip supply remains short, the Samsung Galaxy 8 and the Xiaomi MI-6S phones will not reach volume production scheduled for March and April.  Both use the Qualcomm 835. Telstra in Australia and SK in Korea are ready to go as soon as the handsets ship. Sprint and T-Mobile are racing to be first in the U.S. in 2017. Huawei says over 50 telcos are working with them to plan upgrades, although I doubt many of them will go wide in 2017.

Here's how it's doneFor non-engineers. How LTE gets to the gigabit: 4x4 MIMO, 4 Band Carrier Aggregation, 256 QAM

A near gigabit or better, shared, will be delivered by a combination of more spectrum (three and four band carrier aggregation,) more antennas (4x4 MIMO,) and a 30% improvement in coding (256 QAM.) Each of these techniques is already in production at several telcos; combine them effectively and peak speeds will reach to nearly a gigabit or more. The tower equipment I believe is ready; the phones are causing the delay. Note that the Qualcomm 835 performance is less than those attributes should produce.

Tech publications such as EE Times have been reporting that the tools have been deploying in 2016 in many parts of the globe, from T-Mobile in the U.S. to True in Thailand. When the reporters check with experts, they will find agreement the technology will work. The gig is shared, like all mobile, and the first versions make some compromises. Qualcomm's Sherif Hanna expects average throughputs between ~112 Mbps to ~307 Mbps. Speeds could be as high as ~533 Mbps for 90th percentile users. 

Techtimes reports, "Based on the data indicated in the GFXBench test, the Snapdragon 835 trounced all existing premium handsets including the iPhone 7 Plus, Samsung Galaxy S7 edge, and the OnePlus 3T. …In real world performance, the Snapdragon 835 will purportedly offer 30 percent improvement in terms of efficiency, a 27 percent boost in performance and the reduction of power consumption by up much as 40 percent." Those figures relate to the current networks; on a gigabit network the connection increase from the 256 QAM alone should be 25% to 30%, if well positioned.  

Do remember that the edge of the cell and some people indoors will get real speeds 90%-95% lower. But the takeaway is that real consumer speeds of hundreds of megabits will be practical for good connections. 

 British Telecom is already testing a 2 gigabit system that Huawei hopes to deliver to London's Tech City by the end of 2017. The two gigabit systems will extend the one gig to more spectrum with MIMO working on all bands. They are like the one gig system from Qualcomm with more of same.

My first draft of this story was titled, "1 Gig wireless misses 2016 target on Qualcomm, Samsung delays." Qualcomm in February said equipment would be on the market later in 2016 but it's mostly 6-12 months late. They confirmed the 2016 target in the fall. Telstra in Australia and SK in Korea promised "before the end of 2016" for gigabit LTE. Qualcomm promised their X16 chip would be ready in plenty of time. Samsung promised their "10 nm" fab would be ready for Qualcomm's chip to go into production,

Digitimes reports, "Yield rates for Samsung's 10nm process technology have been low prompting Qualcomm to turn cautious about its product roadmap for 2017, the sources said. Qualcomm originally planned for the Snapdragon 835 and other chips including the 660 (codenamed 8976 Plus) built using Samsung's 10nm process, but has revised its roadmap by having only the 835-series made using the newer node technology."

Digitimes thought TSMC, the largest foundry, also was struggling with 10 nanometer. However, their Elizabeth Sun told EE Times, "TSMC’s 10nm process is totally on track.” 

 

 

5GW News

dave rightStarting July 2017, all new articles are at wirelessone.news. The pols and the marketing folks have made "5G" a meaningless term so the name doesn't fit reporting on advanced wireless. This will stay so old links work. I've reported broadband since 1999 and now is the time for gigabit wireless. Catch a mistake or have news? Email me please. Dave Burstein