Qualcomm today is introducing the new Snapdragon 205 SoC which is an amazing device. SemiAccurate thinks most onlookers will ignore this ‘low low end’ part but it is quite important for carriers and consumers.
On paper the new Snapdragon 205 is nothing more than an SoC aimed at flip/feature phones, you know those things without the performance of a smartphone. The 205 is pin compatible with the smartphone oriented 210/212. For specs it has a dual core ARM A7 running at up to 1.1GHz, an Adreno 304 GPU, GPS and it’s relatives, 802.11a/b/g/n Wi-Fi, Bluetooth 4.1, USB2.0, LPDDR2/3 support, and eMMC4.5. It can playback 720p/30 video and encode 480p/30 in h.264 from a 3MP rear camera or a VGA front camera.
If this sounds underwhelming go look back to the first Samsung Galaxy S smartphone from 2010, it used a single core Hummingbird based CPU at 1GHz. This new ‘feature’ phone SoC would handily annihilate it on most benchmarks. Time does move on even if the Snapdragon 205 is built on a 28nm process. It does sound underwhelming though based on pure specs doesn’t it?
That said it is an important device especially when looking at market impact. Why? Notice we didn’t say anything about the modem on it? In this case the 205 has an X5 modem, that is an LTE Cat4 device capable of 150/50Mbps download/upload. This makes it a killer part for the intended market, carriers will kill to have phones based on the 205.
Why? Spectrum efficiency. Most feature phone SoCs have one thing in common, they are cheap. Really cheap. In fact the first three of the key features list are cost, cost, and cost according to every player selling devices into the emerging markets SemiAccurate has talked to. This is not a small number of OEMs mind you, and most differ on importance of features after that, but the majority of them have at least three more costs in the top 10. Performance, reliability, and modem are usually not in that top 10, things that users care about enough to make the features list on the side of the box are in there. After cost.
And the Snapdragon 205 is almost assuredly going to cost more than the competition if there is any at the moment. SemiAccurate has not heard of any LTE bearing feature phone SoC but there will probably be a few in the near future. If history is any guide, the Qualcomm parts won’t be the cheapest on the market but they will be the best. Again this doesn’t sound like a winning sell to OEMs hell-bent on cost over all for emerging markets.
That is where the 205’s features are a killer sales point, specifically the modem. It is pretty obvious Qualcomm took a bunch of off the shelf parts and played Lego to result in a high-end feature phone SoC with a massively overpowered Cat4 LTE modem. On its own there would be roughly zero chance it would make headway, in the far sub-$50 retail market, a few cents in added BoM can kill a device. Qualcomm is likely to sell the 205 for a few cents more than the craptacular 7th-tier vendors now servicing this market.
So again, why are we saying this is a killer part? Because carriers have limited spectrum and 2G/3G devices take vastly more spectrum resources to accomplish the same tasks as an LTE modem. A carrier can support X users in Y spectrum on 2G. To make the same calls on 3G, it can support a multiple of X user in the same Y spectrum. If you are looking at data, the same holds true but the disparity becomes even greater. LTE can transmit a large multiple of the MB with the same spectrum allocation in the same time as a 3G modem. The higher categories of LTE are likewise much more efficient than the lower categories for similar reasons, mainly modulation.
So for the carriers an LTE feature phone is like gold, it allows them to service more customers with the spectrum they already have. As it was with GbLTE, this higher efficiency means not just more users can be serviced but they can be serviced for less costs. Assuming the carrier is sane they will drop costs a bit and take a bit more profit, plus add more users. It is a win/win/win for them. For users they get potentially lower usage fees, better service, fewer dropped calls, fewer dead air seconds waiting for a time slice, and all the other cumulative benefits. This may not seem like a big deal in 1st world areas but in high density environments, especially those in 3rd world countries, this is a very big deal.
Carriers drool for this kind of thing. It wouldn’t surprise SemiAccurate if they were subsidizing the 205 to get devices with it on the market. The faster LTE takes over the quicker they can de-allocate 2G and 3G resources. That freed up spectrum becomes LTE or more likely 5G spectrum and the virtuous efficiency cycle continues. If you think we are joking about the drooling bit, it is probably an ill-advised term, kill for a fraction of the benefits is perhaps more apropos. If you doubt it, go talk to a few of them. Really, your eyes will open pretty wide when you hear what they say.
For the users they get a vastly superior set of wireless resources, bluetooth, Wi-Fi, GPS, and all the rest that likely come bundled in the Qualcomm modem IP blocks. Along for the ride is VoLTE, VoWi-Fi, and most other modern comms protocols plus dual SIM support. They also get a solid camera with video playback capabilities that are relatively modern. The 205 is a huge step up for the feature phone set, buyers in the target market will probably be quite impressed.
But the killer app for them is much more subtle, security. Remember security and encryption on 3G? Remember when it was trivially cracked? Remember that 2G has about the same level of security as a ROT-13 encrypted file and the cryptographic sophistication of a breakfast cereal prize decoder ring? If you have any idea of how the 3rd world works for payments and micropayments, not to mention attestation and authentication, you know 3G is laughably insecure and 2G is right out.
LTE on the other hand, while not perfect, is far and away better. A secure transmission protocol is nothing without a secure source for attestations and authentication. The Snapdragon 205 looks to change this in a big way, it should have all the latest, or at least new enough for the purpose at hand, crypto infrastructure to keep the financial institutions happy. In short they are more likely to accept a transaction from a Snapdragon 205 than another feature phone SoC. Again think pressure from the carriers and financial side of the world pushing this chip, not to mention governments who are invested heavily in biometrics and security. India comes to mind here.
So we have a second killer app in security but unlike the LTE modem, this one directly impacts users and infrastructure vendors in ways that are noticeable. “This phone is accepted by your bank or similar financial institution” is one heck of a selling point in some areas. Subsidies also come to mind here as well because at the current time in many third-world countries, the carrier is your bank. This grew organically in some areas from trading pre-paid minutes as a proxy for cash but as things become more formalized and higher values are shifted, more security will be necessary.
Those in the western world may not understand the magnitude of feature phone sales. In many lower-income countries the category is the majority of sales. By the best numbers we can find at the moment, feature phones represent about 40% of devices sold world-wide. That is a very big TAM even if the margins aren’t so hot. Small margins on a mere few hundred million sales a year is money I would gladly take but your opinion may differ. SemiAccurate thinks this will be very good business for Qualcomm and carrier pushes and potential subsidies will only make it better.
Devices bearing Qualcomm’s Snapdragon 205 SoC will be available in Q2, the chip is available now. The pin compatibility with the 210/212 should mean a low-cost of development for devices, quicker certification, and all the rest. For those of us in the developed world, we may never see a device bearing this chip on the market but hundreds of millions of them are likely to be sold per year. This is good business for Qualcomm and the 205 will likely kick off a wave of high-end LTE feature phone SoCs. While cost may seem to initially be a barrier, carrier pressure from the top for spectral efficiency and payment security from the bottom will likely mean quicker adoption than most skeptics believe. This is a good part for a potentially massive TAM.S|A
Latest posts by Charlie Demerjian (see all)
- More on Intel’s 10nm process problems - Sep 17, 2018
- Intel puts out another 14nm 2020 server platform - Sep 11, 2018
- Why Can’t Intel Supply Enough 14nm Xeons? - Sep 10, 2018
- Intel can’t supply 14nm Xeons, HPE directly recommends AMD Epyc - Sep 7, 2018
- AMD reintroduces the Athlon name with two CPUs - Sep 6, 2018