Discrete 5G Modem Coming to the Market

The MediaTek M70 discrete 5G modem chip is intended for intel OEMs as the customer. Samples are probably already available now as board-level manufacturing should start in early 2021. Intel will be writing the operating-system-level drivers for this 5G modem and my guess is that it might even be for an Intel ARM based PC. There might not be a reason why Purism couldn’t do the same for the Librem 5 or other Purism devices. This is the first discrete 5G modem that I am aware of. It’s on a 7nm manufacturing process which should create a very high efficiency (low current) use.


bring 5G networking to laptops and PCs ?

if the antenna is indoors and the outer walls are made of concrete (iron-grid reinforced cement) then i would be highly skeptical it would even be worth a shot since the firmware will switch to lower frequencies as soon as it ‘figures’ that the signal is not getting through …

I did some research on 5G. 5G is only expected to go roughly 300 feet to up to 1500 feet if unobstructed, whereas 4G towers go roughly 16 kilometers (5G site coverage being only roughly 2% the geographic coverage of a 4G site). It is expected that it will take 15 years for the 5G infrastructure to fully fill out. Unlicensed 5G systems will start out more at corporate campuses and shopping malls at first, before moving also in to smaller businesses and maybe in to some neighborhoods eventually. Cable companies with large coverage areas can now become phone carriers with few if any barriers. By the time 5G is done filling out, private unlicensed systems should be so prevalent that licensed carriers may find it difficult to compete against the unlicensed 5G systems. 5g is supposed to deliver more data to your device than you can get from your cable carrier directly from the cable modem via a CAT6 cable now, and may even rival today’s optical fiber. Even today’s existing wired/cable/fiber infrastructure in big cities would not support the delivery of that much data to many average users at the same time. 5G will almost be like a glorified WiFi that uses LTE protocols and much larger bandwidths in to the user’s device. Already, in some supported markets, you can’t get a top of the line phone from 5G supported carriers except for phones that support 5G service. But most markets haven’t implemented 5G yet and those markets that have implemented 5G currently have very small 5G coverage areas (some of which are private and used more for testing/experimental use only). However there are huge investments in place that will change that landscape pretty quickly in the near future. The Amateur radio 9cm band (3.3 GHz to 3.5 GHz) has recently been taken away from the Amateur Radio community so that it can be auctioned off to commercial 5G carriers. Currently, the FCC is exploring how to find and re-allocate any possible existing rf spectrum space for 5G use. The reason for this is that the central shared database that all 5G users and operators will share in real time, will make 5G provide the most efficient use of the rf spectrum space possible. But currently, many frequency bands go nearly unused. The re-allocation of spectrum space when over-the-air TV transitioned from analog to digital, left large chunks of spectrum space unused now. So it looks like all guns are loaded and the trigger to an explosive 15-year growth process in communications is about to be pulled.

Big money and extremely high-tech interests (multi-billion dollar silicon fabs and accompanying intellectual property just to begin with) seem to own all aspects of this 5G market. By the time I could find a discrete 5G modem to experiment with, the FCC took away the 3.3 GHz to 3.5 GHz band that I had planned to experiment in, as a licensed Amateur Radio experimenter who might want to learn the GNU Radio software. The same will probably happen to the 5cm, 3cm, 1.2cm Amateur radio bands and up, as modems that can go there start to reach the commercial market. So the only protocol experimentation that appears to be possible and legal, is in the development of commercial products for the 5G market. Any personal interest one may take in to any part of the RF spectrum outside of a firmly protected band, are likely to be swallowed-up by commercial interests before you can even discover them and start your experiments. So unless you want to get in to ‘contesting’ or some obscure experimentation like moon-bounce or metior-scatter propagation, Amateur radio experimentation appears to be dead.


That’s the millimetre wavelength part of 5G and only one part of the spectrum allocated to it. It is perfectly capable of using the same wavelengths as current mobile phone frequencies and will do as the current standards get retired. The other part of 5G, the improved encoding, communication protocols and underlying network structure, is very much universal.

A 5G cell tower operating on 900 MHz will have pretty much the same range as a GSM, UMTS or LTE tower on the same frequency, but offer a little bit more bandwidth than LTE on that frequency.

You’re confusing a small part of the 5G standard (the high frequency signals, which get the bite-size slogan “mmWave” or “NR”) with the whole. 5G as a whole is nice. 5G mmWave is of questionable value precisely because of your concerns (and also because the increased cell tower density makes locating you more accurate).

And yes, current 5G modems are kinda crappy. But that doesn’t mean that they’ll suck in 10 years time. If only there was a phone which contains a user-replaceable cellular modem…


I have to confess that 5G is something I care absolutely zero about (unless there is a forced migration to it at some point). I use mobile data for email, mms, navigation, and an occasional 60-second check of something on the web…hardly things that would require such capacity. I also doubt that I will ever get on board with connected cars and such.

Could someone here comment on the privacy/surveillance implications that have been brought up in the media? Is it something to be concerned about?


The question is whether anyone can make a 5G modem that requires so little current that it can be cooled with just a passive heat spreader on a M.2 3042 card or whether the industry will figure out how to do some new cooling technique like sticking vapor chamber cooling on an M.2 card. From what I have read, 5G modems currently consume between 2.5 and 2.0 times more energy than 4G modems, so it really is a challenge to cool them. Another question is whether anyone will bother with 5G M.2 cards since laptops in the future will become increasingly like smartphones, with everything integrated into the SoC or soldered to the motherboard.

Sadly, big money interests control RF allocation in the US. Getting rid of Ajit Pai as FCC Chairman will help. Let’s hope for another FCC Chairman like Micheal Copps, who at least tried to balance the public interest with corporate interests. I tend to think that Biden will pick people who side with Google, Facebook, etc. over Comcast, Verizon and AT&T, but I don’t know where amateur radio operators will fit into that equation.

mmWave 5G (>24GHz) seems like a huge waste of resources to me, so maybe I should be glad that the US has decided to offer commercial sub-6GHz 5G, since that is the only way that 5G makes much sense to me. Thankfully sub-6GHz 5G can share the spectrum with 4G, so I’m hoping that 4G networks will stick around for the next 2 decades. In my opinion, the industry should have just stuck with upgrading to LTE-A and moved to LTE-TDD bands in congested areas, but that doesn’t have the marketing potential of “5G” and the ability to make people buy new phones.


Very good information amosbatto. I agree.

1 Like

I would assume that you only need cooling when downloading at the highest possible speed or something like that !? So, you could possibly limit the consumed energy indirectly.


While currently 5G networks are still being deployed around the world, and no 5G modems are yet available for the Librem 5, if a 5G modem does becomes available for the Librem 5, you’ll be able to upgrade it yourself without having to buy a new phone. [Edit: It turns out the details behind this are more complicated than I first thought–due to antennas tuned to particular frequencies, a 5G modem inside a Librem 5 would likely not be able to access the new 5G frequencies].

1 Like

Probably from a customer perspective you only care for more bandwidth if you are consuming video content.

However from a carrier perspective, they care to be able to serve more customers per base station etc., so more available bandwidth is a good thing. (That isn’t specifically referring to 5G. The same logic applies at every generation. More available bandwidth can arise from a range of techniques, not just going to ever higher frequencies, which, as has been pointed out, reduces coverage.)

The additional question for the L5 is regarding the antenna. Even if a suitable M.2 card exists and it doesn’t suck the battery dry too quickly and it doesn’t overheat, you still need a suitable antenna.

1 Like

Makes sense.

You need new antennas for mmWave 5G (>24GHz), but as far as I know, the current antenna in the Librem 5 should work for sub-6GHz 5G. A Google search didn’t turn up anything about the need for a different antenna for sub-6GHz 5G, so I assume the standard LTE antenna will work.

1 Like

While the ad-funded media’s general attitude would tend to overlook it, a surveillance concern is a logical consequence of increased bandwidth. With mmwave, it becomes technically feasible to handle the transmissions from surveillance cameras placed on every building.
There are a few obstacles to total corporate dragnet surveillance IRL (camera power, competing surveillance corporations, businesses with their own cameras not needing them right now) but those all go away if a government contract goes out.

But more recently, it looks like Oppo (oneplus) has figured out how to put a 5g-derived mmwave radar precise enough to tell how you’re breathing into a phone. I really hope that what they used was truly a special component and not software analysis of signals from the regular 5g antennas.

Also, the requirement for more towers means more granular triangulation is conceivably possible.

1 Like

Interesting… Thanks!

Even without triangulation, with such small cells, more precise location via the cellular network becomes possible.

I don’t know how high on the RF spectrum you can go and still have usable spectrum for terrestrial phone or data communications. Above 300 GHz isn’t even regulated because it doesn’t appear to have value for any wireless communications. The propagation properties there are probably more similar to visible light (even though 300 GHz is quite a way below the frequency of visible light), than what we think of radio waves as being like. I think that there is a fundamental frequency and several harmonics that tend to attenuate in signal strength significantly when they hit atmospheric water in any form, which is what limits WiFi distance (by design). But the higher you go in frequency, the bigger the amount of bandwidth increases exponentially for each given band. For example, the MF band between 300 KHz and 3 MHz (example: The entire AM radio band between 540 KHz and 1.6 MHz), is significantly smaller than one 6 MHz analog television channel, whereever that TV channel may be on the more expansive spectrum scale anywhere between DC and daylight. The point is that with some of the 5G planned to go up to 100 GHz, we’re going to be having radio waves that are extremely line of sight and limited in distance, almost but not quite like light, in characteristics. But the amount space there is relatively unlimited compared to today’s terrestrial communications systems. The sub-6GHz 5G communications systems will have some spectrum space limitations but will be more like radio waves as we know them in existing 4G systems. Satellites have altitude line-of-sight limitations above 24 GHz (for example). So there is no problem at 24 GHz for satellites. But how high on the dial (spectrum) we might be able to go for practical cell phone use with many many cell sites (of sorts) has yet to be seen. Radio wave propagation characteristics over various frequencies is a mammoth area of study.

In the higher frequencies the 5G antenna’s (phased array) are highly directional. The antenna bundles energy in the direction of the communicating device and can pinpoint the device in 3D. Based on ships radar in the same freq. band I expect an accuracy of ca. 1 m - 2m in either direction (500m distance).


Please have a look at my quote of Kyle on antennas above:

New Radio (NR) operating bands are listed here: https://www.sqimway.com/nr_band.php, just as overview. @StevenR, perhaps, if I’m getting this thread right one simple (simplifying purposely) example would be this S6502 modern smartphone (to this topic related) with MT6873 (see bottom of the linked page) offering following Bands:
2G: 900/1800 MHz
3G: 900/2100 MHz
4G: B1/B3/B8
5G: N1/3/8/28/41

As this 3GPP 5G-NR Overview (.pdf) from Dec 2017 explains: “Note: Existing LTE operating bands 1, 3, 8, 28, 41, and 74 (related to Japan) are also going to be introduced as NR operating bands as n1 (FDD), n3 (FDD), n8 (FDD), n28 (FDD), n41 (TDD), and n74 (FDD).” Also, somewhere else, I saw that only the primary SIM card slot accepts/works with this NR Bands: n1 / 3 / 28 (UL 703 - 733 MHz, DL 758 - 788 MHz) / 38 / 41 (2.496 ~ 2.690 MHz) / 77 / 78 / 79.

I work for a telco and the issue is always based on a business case. In low density areas, rural areas, millimetre wave makes no sense when you consider the cost of each tower, and the density of towers you require with that technology. Urban centres sure, you can get more granular with your infill sites to reach tricky spots. No matter the tech, it’s all about the frequency as someone else mentioned. 700 and 900 MHz has much better range and penetration than these 25 GHz+ systems. And honestly nobody is really sure what the health implications of being exposed to such high frequency radiation is.


I don’t know if this is truth or urban legend. But I have heard that people are starting to find dead birds at the base of some 5G towers.