Do not buy 5G smartphones - at least in the near future

Analysis of the many ways in which the first generation of equipment with 5G support (temporarily?) Will destroy existing developments in the design of phones

"5G is here"
5G is already here, but this does not mean that you have to buy it.

2019 will be the year of 5G - at least, such statements continue to be made by the cellular industry. We will see the launch of sales of several smartphones with 5G support from such OEMs as Samsung, Motorola and OnePlus, and operators will get out of their skin telling you how cool their new 5G networks are, despite a bunch of stars after the advertising text. I wanted to come up with some kind of joke that demonstrates the absurd level of hype around 5G, but it is very difficult to outdo real quotes from business directors - for example, Verizon says 5G will "dramatically improve our global community." Faster mobile internet is already on the way, but is it really time to worry about it?

Recently, Qualcomm made a big announcement about its chips for 2019, and since it is the largest supplier of chips for smartphones in the world, we can get a pretty good idea of ​​what the future hardware for 5G will look like. The industry is struggling to present 5G as the “next hit”, but the 5G of the coming year will definitely be the first generation breakdown. The very first 5G users will have to accept the full range of trade-offs. And due to the fact that where you live, in principle, there may not be 5G coverage, it may be better to just wait a year or two.

5G mmWave Reference: Using Spectrum That No One Needed

5G is a brief designation of next-generation cellular network technology, launched in 2019. This whole numbering system with the letter G began in the 1990s with the launch of GSM technology, which was called the “second generation”, or 2G mobile network technology. GSM updated the very first networks from analog to digital, and therefore old analog networks retroactively called 1G. Since then, we get all the new G after each update of the networks about once every 10 years. These stages of development gave us such important things as SMS and MMS messages, IP-based networks and mobile Internet, and, of course, even more speed.

Today, modern smartphones work on 4G LTE at frequencies in the range from 450 MHz to 5.9 GHz. The transition to 5G will include improvements to the existing LTE infrastructure, but the defining characteristics of 5G is the addition of a new piece of spectrum in the range from 24 GHz to 90 GHz. The industry has agreed to call this range mmWave (from "millimeter waves"), and they will require the installation of new hardware in the telephone, new equipment on the towers and major changes in the telephone and telephone network operation patterns.


mmWave offers a wide spectrum bandwidth, but is difficult to use.

We are used to the fact that these updates of the G network are accompanied by convincing presentations about how much better everything will be after implementation, but the transition to 5G mmWave cannot be called a convincing argument. Since the frequency of the mmWave range is much higher than that of LTE, it lacks the disadvantages. The range has a smaller reception distance, it penetrates worse through obstacles compared to LTE. The mmWave signal is blocked by buildings, trees, and even by hand. mmWave does not work well during rain or fog, and at 60 GHz the signal is absorbed by oxygen. Yes, yes - part of the mmWave spectrum is blocked by air.

Given the number of obstacles that need to be overcome, mmWave seems like a terrible segment of the spectrum to build a mobile network on its basis, until you consider two key points: increasing the frequency means that mmWave has a high bandwidth and a short waiting time if everything works out like it is necessary, and, most importantly, this spectrum is free. mmWave is not used for almost anything, since working with it is one headache. Therefore, if you can solve all the problems of implementation, you suddenly have a huge workspace. And this is the first thing that companies talk about when the mmWave conversation begins. They say that this whole undertaking will be extremely difficult and very difficult, but worth it.

LTE appeared in 2011, and in recent years there has been significant progress in reducing, accelerating and improving the performance of 4G smartphones. With 5G we will lose quite a large part of this technical maturity, stuffing a bunch of new expensive equipment for 5G into devices.

Separate 5G modems - more components, more consumption, less battery

Today, almost all smartphones work on the basis of a single chip, which is appropriately called a “ system on a chip ” [System on a Chip] or SoC. The name tells us that these are the main components with which you can create a computer on one tiny chip. Usually there is a bunch of CPU cores, GPUs, ISPs for the camera, WiFi and more. The memory is not technically located on this chip, but to save space, it is placed on top of the SoC. The main component, which is not on the SoC, is the storage, but also tiny chips that control power, audio, Bluetooth, NFC and everything else are usually scattered around the motherboard. And then the task of the motherboard is already becoming to connect everything with everything and quickly get out of the way, because as much as possible of the phone must be filled with battery.

The point is that the place inside the smartphone is very valuable, and although we can not do anything with the size of components such as the SoC, camera, SIM-card or USB-port, we can change the size of the battery. That is, “smartphone size” means “battery size”. If something grows in size, the battery is reduced. Anything that requires additional components reduces the battery. The battery gets the remaining space of the smartphone (this is the argument against the headphone jack).

In recent years, smartphone manufacturers have tried to convince us that we don’t need a headphone jack, and that removing them will reduce the complexity of the system and make room for the battery. Razer Director Min-Liang Tan even rated it quantitatively : he said that by getting rid of the headphone jack on the Razer Phone, the company could increase the battery capacity by 50 mAh.

And why is this important in the article about 5G? In short, 5G mmWave will require much more iron than 4G, which raises all these problems with the size of the battery and the complexity of the device.


For 5G, a separate chip is needed, even with the new SoC from Qualcomm

The biggest advantage of Qualcomm in the 4G era was their modems. The combination of its own technology and intellectual property rights has allowed the company to become the only chip manufacturer capable of combining SoC and modem on a single chip and selling it all over the world for little money.

Qualcomm is the only company with an affordable SoC + modem solution that has a significant worldwide distribution network.

Samsung offers integrated modems in the Exynos line, but in the US, China, Latin America and Japan, Samsung devices still use Qualcomm chips. Huawei and MediaTek are in a similar position: they can make LTE on a chip, but they do not distribute these chips around the world. Apple does SoC for the iPhone, but uses a separate chip for the modem. Apple left Qualcomm modems on the iPhone XS in favor of Intel modems, but this gave rise to a whole bunch of litigation between Apple and Qualcomm.

Qualcomm owns cellular patents in some countries, and it aggressively uses its intellectual property to maintain a dominant position. The company has already faced numerous legal proceedings from many companies in many countries around the world in connection with unfair patent licensing, but this did not affect the reality of the market, in which everyone except Apple uses Qualcomm modems.

The solution on a single chip is a great advantage, allowing you to make the motherboard smaller, cheaper and simpler, leaving more space for the battery. If you bring everything together on a single chip, this will save energy during operation, since usually one chip consumes less energy than two. For years, Qualcomm users enjoyed using SoC with 4G LTE modems, and the company used this advantage to dominate the market. Today, it practically holds a monopoly among high-end SoC manufacturers, since almost all Android-based flagships use Qualcomm's SoC.

Qualcomm has recently demonstrated its flagship SoC for 2019, Snapdragon 855. And although the company is spending a lot of effort to promote the hype around the 5G Snapdragon 855 capabilities, it will not have a 5G mmWave modem on board. It will have built-in support for LTE, as usual, but phones will need a separate 5G modem - Qualcomm will lose its advantage of a single chip when 5G appears. As I explained above, this will mean a decrease in battery and an increase in power consumption.

We have already experienced the whole story with the "first generation network equipment." When the transition to 4G began, the first batch of equipment also came with separate modems - we will see this compromise in the case of 5G. The most famous example was the HTC Thunderbolt, the first 4G device on the Verizon network. He used the Qualcomm Snapdragon MSM8655 SoC (before simplified model numbering) with a separate Qualcomm MDM9600 LTE modem. And this Thunderbolt was a disaster, since all its new 4G equipment was bundled with a battery of only 1400 mAh. The phone was slow, hot, buggy, and quickly spent the battery. Thunderbolt is constantly on the list of the worst phones of all time, and one employee HTC even apologized for its creation. New network equipment can turn into a disaster if it is applied incorrectly.

First-generation 5G giant chips from Qualcomm

I can not guarantee that the first generation of 5G equipment will be as bad as Thunderbolt - since then, much has changed in the design of smartphones - but concern about early versions of network equipment remains. New network equipment is new because it is being reduced for the first time enough for it to fit in a smartphone. Companies do not wait until it is easy for them to meet size requirements in order to release a product; they start developing a new phone at the same second when the network equipment barely fits. These 2019 devices will most likely be filled with iron for 5G - the largest of the existing iron for 5G.

Let's compare the internal components needed to make 4G and 5G networks work. With regard to the main chips from Qualcomm in 2019, a Snapdragon 855 will be installed on a 4G phone - that's all. An LTE modem is built into it, so a separate chip is not needed, and antennas for LTE are tiny wires, usually integrated into the motherboard. 5G is a completely different thing: you will need a Snapdragon 855, a Snapdragon X50 5G modem, a set of antenna modules QTM052 - and these are real chips, and not just wires or motherboard tracks.


Qualcomm chips near cents for scale

I have not seen Qualcomm publications with 5G or Snapdragon 855 chip sizes, but they habitually post their photos next to cents. They should say, “look at how small our chips are,” but in addition, they give an excellent scale - we can load photos into Photoshop and get images of chips in scale.

It turns out that the 5G chips of the first generation will be hefty - at least compared to other extremely small components. The X50 modem and the only RF module take up more space than the entire Snapdragon 855. Keep in mind that the 855th is completely SoC, that is, everything you need for your computer, plus the built-in 4G modem, so it's unbelievable that the 5G modem and one module RFs take up almost as much space as all other key components, with the exception of storage. So far, 4G has a neat layout on the chip, and 5G will have to deal with massive chips.

"Exponentially" more complex iron and a lot of difficulties in developing

At the Qualcomm Snapdragon Summit conference in Hawaii, held in December, company president Cristiano Amon talked a little about how difficult it is to create a smartphone for 5G. On one of the presentation slides, it was mentioned that 5G “exponentially” complicates the development of smartphones. And although Amon had hoped that this would be an impressive compliment to the company's engineers, I think it sounds like horror killing a battery. Difficulty is bad. Difficulty is expensive.

The point is not only that iron for 5G is larger than iron for 4G - you will need to fill the phone with iron for 5G to make it work. mmWave penetrates obstacles so badly that the signal can be easily blocked by hand - and for a device that you keep on using, this is a problem. To avoid a situation in which you " hold it wrong ", Qualcomm decided to fill the phone with several 5G antennas.









So your phone will require not only the Snapdragon 855, the X50 modem and the QTM052 antenna module - it will need several QTM052 chips. The company's specifications state that to bypass the mmWave restrictions, four modules will be needed, placed in different places, although more advanced versions of the scheme can do without three. All diagrams for 5G depict four antenna modules — one on each side. The reference design of the 5G prototype shows three 5G antennas - top, left, and right. Motorola's 5G Moto Mod is the closest to the real consumer product that we have today - four antennas: left, right, and two at the top.

And no matter what configuration OEMs choose, it appears that all phones will have several antennas, and the phone will be able to choose an antenna that your hand is not blocking at the moment. For example, if you hold it in portrait mode, then your hand can block the left and right antennas, and the phone can switch to the top and bottom. If you use a landscape arrangement, everything happens the other way around. 5G will independently switch to antennas capable of working with the signal.

More antennas - more complex interiors and less free space. Returning to the photos of the chips in scale, we see that a set of chips will occupy a huge amount compared to 4G. I don’t know what their thickness is, but if measured in pixels, then 5G takes 3.3 times more area than 4G on SoC. For an industry that can no longer provide a headphone jack, since it takes up too much space, all this additional 5G equipment is hard to justify.


Antenna side? A typical smartphone has a metal frame around the perimeter - and this option is clearly not going to work.

But schema development becomes more complicated than simply placing all the components on a flat sheet. Several antennas are needed to implement a technique called "three-dimensional beam formation" [3D beamforming]. Instead of sending a signal in the form of a wave (imagine a WiFi symbol), a beamforming antenna can detect a station and trigger a concentrated signal directly into it. In this case, the station will detect your phone and send the beam directly to you. Qualcomm says this technology is a key feature of 5G devices, as a concentrated beam will help overcome some of the mmWave problems associated with distance and penetration. And for now, all the Qualcomm specifications and all the hardware say that these antennas are located perpendicular to the motherboard, on the edge of the phone.

If for 5G electronics it is really necessary that a large surface area of ​​the antenna is on the side - and so far all the evidence confirms this - this raises some interesting thoughts about the design. Today, wireless charging has led phones to move from all-metal cases to glass covers, allowing radio signals to pass through the back of the case. Since glass is not such a strong material, these phones have a metal frame, reinforcing the structure, which is usually visible from the sides. It is unlikely that a smartphone with a metal rim will be compatible with 5G antennas located on the sides. Of course, the glass rim also can not be, and what remains - plastic? 5G can open the era of smartphones with plastic sides. I see no other way to make these side antennas work.

In addition, someone will have to pay for all these additional components of 5G, and, apparently, it will be done by the consumer. OnePlus director Pete Lau recently told The Verge that 5G will make phones more expensive by $ 200- $ 300. And this is a lot of money for a phone that will be larger and work less than a 4G phone.

Violent reality of 5G release

5G will obviously require smartphones to make compromises in the field of iron - at least in the near future. Switching to 5G will mean accepting many compromises compared to a 4G phone - and what will we get in return? In 2019, only “possibly faster internet”, which depends on a whole heap of variables, mostly related to your location. First, you have to be in a city where 5G works, then you have to be in a certain place where you can receive a 5G signal, and then you have to decide whether you need this increase in speed at all.


Judging from the diagram, Qualcomm imagines that mmWave will be used only in cities, due to distance limitations. In other places there will be only LTE of various classes.

The high speed on 5G is connected with the ability to catch the 5G signal in the first place, but for now the implementation of this technology seems like a nightmare. Since 5G operates at higher frequencies and penetrates obstacles much worse than 4G, business will not end with updating base stations. Outside the buildings, an inelegant solution to problems with the range of work and weak penetration of 5G will be to “build more towers,” that is, slow negotiations over property rights and placement of towers.

Because of the limited range so far it is not even worth such a task, cover 5G mmWave all around. The director of Qualcomm described the 5G architecture in the USA as using mmWave in “dense urban development” - and only LTE will remain in other places. Using mmWave in cities is quite reasonable, given the problems with range. The construction of a huge number of new cellular towers for 5G will not pay off outside the city, so LTE will have to be content with it. And even if you find yourself in a city with a 5G coating, it will be difficult to catch him indoors because of problems with penetration through obstacles. Most likely, in order to catch 5G in buildings, you will need to place the towers inside them (femtocell access points), or you will roll back to LTE. No industry representative declares rejection of LTE, and moreover, the current plan implies an improvement in LTE in parallel with the introduction of 5G.

So, despite the big hype about speeds achievable at 5G in principle, the real modems of 2019 will not get a tenfold increase in speed, which is so often mentioned - especially when LTE improvements are made. The 4G LTE modem onboard the Snapdragon 855 - and this solution on a single chip without any compromise inherent in 5G - received a new theoretical maximum speed of 2 Gbit / s, according to Qualcomm. The X50 5G modem and a handful of radio modules can more than double this theoretical download speed, reaching 5 Gbit / s.

But both of these numbers are theoretical speeds that can never be achieved in reality. However, keep in mind that for perfect viewing of high-quality video with 1080p resolution at 60 frames per second, you only need a speed of 0.006 Gbit / s. Even insane things like 4K streaming video require only about 0.025 Gbps, so 4G current speeds are enough for everyone. I am sure that sometime when we start streaming data for virtual reality to a computer in a helmet, we can use fast Internet 5G. But for my next LTE smartphone - which has not yet accelerated - should be more than enough.

Let's return to the topic in 2020

The 2019 5G mmWave will give us thicker, hotter and more sophisticated phones that use more power and cost more. So far we have no commercially available examples, so the real consequences of all these deteriorations have yet to be understood, but it cannot be denied that the first-generation 5G hardware will not reach the more mature 4G schemes. And since the 5G networks are still in the bud, and for a phone with 5G they are going to ask for an additional $ 200- $ 300, it seems that for the user this option will not be justified.


Qualcomm is right about the tip of the iceberg, but buying only the tip is not worth it.

No technology appears fully formed. These early 5G designs need to be carried out, they lay the foundation for future improvements. In the future, 5G technology will be important, at least for large cities, and is needed to help mobile networks survive the constant demands for increased bandwidth. But this does not mean that you are obliged to spend money on 5G today.

After a couple of years, the iron for mmWave will become more mature and integrated. One of the signs to watch out for is whether Qualcomm’s SoC will include an integrated 5G modem in 2020, since this will bring back the many advantages of the system on a single chip that 4G now has. But the immaturity of 5G will not stop the oncoming advertising storm. Soon you will see the launch of the first 5G phone from operator X, and all operators of the planet will advertise 5G as the best thing that has happened since the invention of sliced ​​bread. Do not believe the hype. In the near future, if you can buy a 4G version of the phone, then perhaps you should do so.