5G’s Achilles Heel: Heat

5G promises a lot. a mobile internet of things, new immersive VR and AR experiences, lower latency, washboard stomachs. But something the industry isn’t addressing is that the devices themselves heat up. A lot. This from Digits to Dollars‘ Jonathan Goldberg: 

5G phones get hot. Really hot. Probably not hot enough to ignite your battery (probably), but enough to generate a definite burning sensation in your pants pockets. At Mobile World Congress in February, we spoke with an engineer from Sony who was demo’ing a phone (behind glass) that was clocking 1 Gbps speeds. Wow, fast. We asked the engineer why it was not going faster and he said “It overheats.” A good solid answer, from a nuts-and-bolts-and-antenna person. We will wager any amount that at next year’s show, no one on the floor will be as open about this problem.

The industry, Goldberg writes, is tackling this issue by er, ignoring it. And indeed the standard response appears to be that “we’ve seen heat problems with every new generation and what we have with 5G is nothing significant, 3G was way worse,” as one commenter said he’d been told at a 5G conference. But that may be underestimating the problem — Goldberg says the “heat budget” is 67% higher than current phones. (Heat budget is the total amount of thermal energy transferred to the chip when the device is in operation.) And he points out that both no-one seems yet to be offering a solutions and “solving the issue in 3G broke a couple vendors.” 

Some background: what we call 5G is actually two stages of technology. What most carriers are currently rolling out is phase 1, or what is called, confusingly, sub 6, an evolution of 4G that bring (quoting Goldberg again in a different post): “modest improvements in data rates as well as some important, but hard to observe, changes in the software the operators use to run their networks.” The big step will be the second phase, mmWave, “will bring much more tangible changes, notably including data rates at or above 1 Gbps.”It’s these mmWave radios that are (indirectly) causing the problem. 

As I understand it, these mmWave operate at very high frequencies — close to microwave — which require high clock speeds in the chips. The heat this creates is concentrated in a small subset of the electronic components within the phone, and there’s no easy way to move that energy around. Goldberg again: 

Of course there are some solutions, but none of them are complete and they all have serious drawbacks. It turns out that the way we cool electronics has not advanced in 40 years. There are really two methods used currently to cool Things down- Fans and Dissipation.

Fans are what you think they are. Anyone who has ever opened up their desktop PC or overclocked their laptop knows what these look like. But fans have two problems: they are big and they have moving parts. Both of those require design decisions that go counter to every mobile design trend in the past 15 years.

Dissipation is just the idea of moving the heat around to hasten air cooling. In a PC, this is typified by those funny looking prong-things that sit on top of CPUs. Those things are too tall to fit inside a 10mm thick phone. So for mobiles, OEMs are looking at using ‘straws’, or copper pipes that span the length of the phone. These take up a lot of space and inserting a large conductive element (copper!) inside a phone wreaks havoc on mobile radios, (i.e. hurting data rates).

We all know the problems of overheating phones, but what is surprising is how little this issue seems to be addressed. Goldberg says that this is a problem on a whole new level to previous generations, and one that is only now being addressed: “The problems with 5G mmWave are larger and will not go away as quickly. Handset makers are just waking up to the existence of this problem.”

The only place to find discussion of this issue appears to be in academia, which itself notes the lack of discussion. In a paper published last year three researchers at the Huazhong University of Science and Technology wrote (PDF):

the heat dissipation of smartphones restricts the maximum receiving rate of smartphones. Although the maximum receiving rate of smartphones is restricted by the computation capability and heat dissipation, detailed studies of basic models used for evaluating the maximum receiving rate of smartphones are surprisingly rare in the available literature.

The researchers ran their own tests and reached some sobering conclusions: 

– anything above 4 Gbps and the temperature of the smartphone reaches above 45 C “within a few seconds.” (5G has promised peak data rates up to 20 Gbps and Qualcomm’s first 5G modem “is designed to achieve up to 5Gbps downlink peak data rate.” So the smartphone has to “decrease the computation capability of the chip to reduce the heat generation, e.g., decrease the working frequency of the chip, to prevent low-temperature burns on the user’s skin. Thus, smartphones cannot sustain the original receiving rate and may even have to shut off wireless communications.” This is obviously not an optimal outcome. This is already happening with the first mmWave 5G rollouts (what AT&T calls 5G+) — which, remember, is not the one that involves mmWave radios: The Wall Street Journal wrote in July that their Galaxy S10’s 5G switched off in the Icelandic summer. Others have reported similar problems.

The researchers recommend that to address this”using new materials or redesigning the components’ structure to improve the heat conduction rate from the chip to other low-temperature components in smartphones. Additionally, mobile edge computing, one of the 5G technologies, can be applied to improve the maximum receiving rate of smartphones by offloading the computation assignments in the chips.” It’s hard to imagine that would be a welcome advance, since as I understand it it would mean transferring a lot of the hard work from the phone to the base station — and who exactly would pay for that? 

The researchers are, in their academic way, somewhat scathing of how the field has failed to address the serious matter of device heat: “In 5G and future 6G cellular networks, most of research is focused on the core networks and BSs. However, many potential impacts triggered by the maximum receiving rate of smartphones have not yet been investigated. How to design reasonable mobile terminals for matching with 5G and future 6G wireless communication systems is still an open issue for industries and academic researchers.” 

That was a year ago. One can only hope the device manufacturers are addressing this. For now, it seems to make sense to take 5G promises with a pinch of salt and a bucket of ice.