The stuff in that article doesn't address carrier-imposed data caps. As a cellular user, I just don't really care about the network that moves my data around unless its broken or is working poorly. Preventing it from being broken is the network provider's problem. It's the provider's problem to efficiently use available frequencies or whatever else. If new technology doesn't change my data cap, then going faster means nothing to me but that I can hit that cap a little more quickly. I still have to sit around fretting about how much data of my plan that I use regardless of how much faster it gets so nope, I have nothing to get excited about.
Ivan currently visiting India and amazed at the data caps on 4g here. 10gb a day for a few dollars per month, I kid you not. Coming from Germany with my 40 bucks for 3gb/month I suddenly feel in the 21st century.
I blame the oligopoly of 4-5 providers at most which have bought up all small ones consolidating their power. We need 30's style anti-monopoly enforcement back if you ask me, the deregulation and small government craze started in the 70's has given us enough crisis, corruption and monopolies.
Being in India and just getting a taste of LTE, I cannot help but wonder: is there even a point in getting to 5G speeds beyond vague definitions of "progress" as a numbers game? Is it really as simple as "the company would have nothing to do otherwise"? I mean, even with the kind of user density we have here, LTE is mighty quick...
The main goal of 5G is to make high-capacity mesh networks, from what I'd read previously. In the near future, you might be able to access certain local resources at high speed, almost like you're on a LAN. In those cases, gigabit speed would be good.
Problem is not speed but data cap. Having 5G is meaningless unless telecomm allows you to have data cap like 200 gb or else you will used up all your data or pay a hefty fee.
Data caps exist because the bandwidth available over lte frequencies is limited. 5G aims to alleviate bandwidth limitations by adding higher frequencies and more efficiently using existing ones.
So no, having 5G is not meaningless. It will be very important as more and more services move onto cellular networks.
I'm kinda with Cliff on this. Even if 5G removes some of the limitations telcos are running into, it doesn't mean they're going to pass their customers more data for the same price. They have to pay to upgrade their networks and expand to deal with more people/devices on those networks. Plus businesses have to show growth or their shareholders aren't happy. If 5G makes things more efficient, the very last thing that'd happen is giving customers more data or the same data for less cost. Like Cliff says, the data caps and not speed are the problem for customers right now.
5G infrastructure definitely won't incentivize telecoms to increase caps immediately, but think back to pre-smart phone days for internet access on mobile devices versus the typical data plan size now (and effective cost in $/MB or $/GB). I agree that short term they will use 5G as a means to widen their margins, but long term the pricing window should shift towards larger caps. The secondary implications of effective mesh networks for diagnostic use and applications like vehicle-to-vehicle communication are probably a more important effect of 5G development.
I tend to agree with this sentiment when it comes to wired networks. Upgrading them is a straight up investment that pays off over time. Wireless is a different case, you are restricted by available spectrum and the spec in use. LTE does not permit you to put as many towers as you want into an area, they must e a minimum distance from each other, and the performance of those towers is restricted by the specification and the frequencies a given carrier owns.
Usage caps make far more sense in the wireless space than they do in the wired arena.
Wired networks suffer from congestion just like wireless networks, even fiber optic cable has limits on the amount of data that can flow through the pipe at any given time, hence the data caps on wired networks that is cropping up these days, Cox Cable implements a 1tb data cap on most plans these days, DSL hasn't yet that I have seen, but their speed on regular DSL is so slow that you would be hard pressed to reach 1tb in a month.
They do, but nothing like the same. You can put Terabits/s down a pair of fibres - you can also add more fibres as close as you like without causing a problem, you can't do that with wireless, the spectrum is the spectrum, you can't add more 5Ghz, the 5Ghz is the 5Ghz you've got and that's it - no matter how much you money you might have for upgrades; if it's congested it's congested.
While I believe from the bottom of my heart that capitalism as a system begets greed and I do agree with you, there might be more to the picture: What if cellular networks here can be monetized to a greater degree on account of a bigger user base, thus allowing higher revenue from (say) the same area? So in the US say you'd have a 100 people using a network and here you'd have say 400. Thus data limits are the only way to get the sort of revenue you want (profit margins could be anything) in a low user base density. Currently I pay INR 400 for 70 days of LTE, allowing 1GB per day of full speed data (and unlimited 128Kbps until midnight).
Are you sure enough people are using LTE and doing heavy load on it? No Carrier has yet solved the user density problem in Major Cities areas, especially in high density region like South East Asia.
Massive MIMO are now being deployed, and for TD-LTE Network that part of the equation is pretty much solved, as it brings 10x capacity improvement with No user equipment upgrade.
Massive MIMO on FD-LTE, where 70% of the world uses, is pretty much useless at this stage.
I've used a very popular LTE carrier in a very very high density place (a suburban train station that's one of the busiest in the world). And I can't even load a webpage there. But this case is actually more extreme than you'd see in more developed countries, right? I've moved back to a town (relatively speaking) and now speeds are all right
You are right, but. Better see companies create a network of 5G tomorrow, than expand a network of 3G or 4Gtoday . Because today is just a numbers game. In 10-15 years, consumers will probably be happy seeing at least a 5G in their connection status.
1) Hello Apple (let's face it, who else uses a discrete baseband nowadays, at least in a smartphone?).
2) Does the 7660 support CDMA? Can't see any mention of this in the release (although it says the 8060 supports it). Was under the impression Intel were bringing in CDMA to displace Qualcomm but maybe it isn't ready yet?
7560 Support CDMA, but given latest report show Apple will still be using Qualcomm for CDMA in their next generation of iPhone, this means CDMA support on Intel's first CDMA likely isn't good enough yet.
So 7660 may be the Baseband that will completely displace Qualcomm in iPhone 2019. Given how Intel put out this announcement rather early I expect the design is pretty impressive.
It’s nice to read about all of this while people around the world are stuck at about 20Mb/s,now, with a very few able to get to 50. Unless you’re in a testing area, and right near the tower, much higher than that is a bust.
20Mb/s?! My wired internets at home are only 10 (thanks bunches DSL)...I wish I could get anything close to 20 on any connection and it's not like I live in the middle of nowhere either. *pouts*
Intel only has one customer for its Baseband and that is Apple. So we know for sure the XMM75 series will be in next iPhone, and presumably the XMM76 Series in 2019 iPhone.
Last time we knew the XMM75 will be Intel's first in house fabbed Baseband, and based on 14nm. So are we going to assume XMM76 being on 10nm?
Another Interesting point is this is the first time Intel has a pre announced Baseband that is better and faster then everyone else has on the Roadmap, that is including Qualcomm and Huawei. Although we are hitting diminishing of returns in terms of real benefits.
That is the first Intel Baseband on iPhone. No one has tested the newer one yet. Although I doubt it will be "better" then Qualcomm, but likely improve to a point where it makes little difference in real world.
Qualcomm had had 5g test networks in place for several months now, everything I read in this article that will be in Intel's 5g modem is already in Qualcomms modem, and there is a typo in the article where they mention a 58ghz band (there is no such thing, as that would be in x-ray territory). Since Qualcomm will be shipping their modem in 2018 and test units are already out to phone manufacturers it would seem that Intel is a little behind. If Intel really wanted to get their feet into the SoC and 5g modem business they would have taken more of a role in developing the 5g standards rather than letting Qualcomm be the major player and getting all the standards based patents
Well Intel is partnering with everybody ( apart from Qualcomm ) to develop the 5G. So the battle on Standard based patents is a lot more complicated, including lots of politics.
There is no defined "5G" yet. So what everyone is talking about right now is mmWave. But before any of mmWave happens, carrier has to implement Massive MIMO and LAA first, as well as many other LTE Advance Pro, 4.9G or 3GPP Rel 14 / 15 before 5G happens.
As a matter of fact I think Carrier will likely brand their 4.9G Network as 5G, and then let 5G evolves from it. Much like what happen with 4G.
All I can say is FINALLY! Qualcomm now has some real competition. Intel is first out of the gate with 5G so they will get their patents filed early. Now Intel needs to work on integrating and licensing out these modems for use in SOC's instead of just offering them as a standalone modem. Integration is important for mobile as precious PCB space is not taken up by the modem.
To little too late, Qualcomm already has test 5g networks in place, has had for several months now and their 5g modem will be incorporated into their next SoC series (not 2 years from now), Qualcomm was the major player in implementing the 5g standards and already patented their standards based 5g technology. Everything that I read in this article is already in Qualcomms 5g modem that will be shipping in 2018. Intel is a newcomer in the mobile communications business and their only customer that really matters is apple, Samsung uses Qualcomm or their own SoC and modems, Huawei uses either their own SoC or Qualcomm, and are designing their own 5g modem so that leaves Intel's market share pretty thin
I really don't get why they bother so much improving maximum speed supported by modems each year when the actual network doesn't even support a quarter of the 1Gbps at best. The industry today has taken the obsessive route of improving things at a very fast and unnecessary rate, just for the sake of better spec (read higher numbers) on a new device so that people will be motivated to buy it. Stupid really...
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
41 Comments
Back to Article
PeachNCream - Monday, November 20, 2017 - link
Data caps make it hard to get excited about faster cellular modems. :(SydneyBlue120d - Monday, November 20, 2017 - link
In fact, you should:https://semiaccurate.com/2017/11/14/qualcomm-shows...
PeachNCream - Monday, November 20, 2017 - link
The stuff in that article doesn't address carrier-imposed data caps. As a cellular user, I just don't really care about the network that moves my data around unless its broken or is working poorly. Preventing it from being broken is the network provider's problem. It's the provider's problem to efficiently use available frequencies or whatever else. If new technology doesn't change my data cap, then going faster means nothing to me but that I can hit that cap a little more quickly. I still have to sit around fretting about how much data of my plan that I use regardless of how much faster it gets so nope, I have nothing to get excited about.jospoortvliet - Wednesday, November 22, 2017 - link
Ivan currently visiting India and amazed at the data caps on 4g here. 10gb a day for a few dollars per month, I kid you not. Coming from Germany with my 40 bucks for 3gb/month I suddenly feel in the 21st century.I blame the oligopoly of 4-5 providers at most which have bought up all small ones consolidating their power. We need 30's style anti-monopoly enforcement back if you ask me, the deregulation and small government craze started in the 70's has given us enough crisis, corruption and monopolies.
ads295 - Monday, November 20, 2017 - link
Being in India and just getting a taste of LTE, I cannot help but wonder: is there even a point in getting to 5G speeds beyond vague definitions of "progress" as a numbers game? Is it really as simple as "the company would have nothing to do otherwise"?I mean, even with the kind of user density we have here, LTE is mighty quick...
mkozakewich - Monday, November 20, 2017 - link
The main goal of 5G is to make high-capacity mesh networks, from what I'd read previously. In the near future, you might be able to access certain local resources at high speed, almost like you're on a LAN. In those cases, gigabit speed would be good.Cliff34 - Monday, November 20, 2017 - link
Problem is not speed but data cap. Having 5G is meaningless unless telecomm allows you to have data cap like 200 gb or else you will used up all your data or pay a hefty fee.saratoga4 - Monday, November 20, 2017 - link
Data caps exist because the bandwidth available over lte frequencies is limited. 5G aims to alleviate bandwidth limitations by adding higher frequencies and more efficiently using existing ones.So no, having 5G is not meaningless. It will be very important as more and more services move onto cellular networks.
PeachNCream - Monday, November 20, 2017 - link
I'm kinda with Cliff on this. Even if 5G removes some of the limitations telcos are running into, it doesn't mean they're going to pass their customers more data for the same price. They have to pay to upgrade their networks and expand to deal with more people/devices on those networks. Plus businesses have to show growth or their shareholders aren't happy. If 5G makes things more efficient, the very last thing that'd happen is giving customers more data or the same data for less cost. Like Cliff says, the data caps and not speed are the problem for customers right now.ads295 - Tuesday, November 21, 2017 - link
What sort of data caps do you guys have on LTE? I get to use 1GB per day of high speed LTE at INR400 for 75 days plan validity.PeachNCream - Tuesday, November 21, 2017 - link
I'm on a no contract phone so I typically get 180MB per months, though I sometimes buy an extra 1GB of pure data for another $10 USD.jospoortvliet - Wednesday, November 22, 2017 - link
3gb/month for 40 euro in Germany. Yes per month. Yes that is about INR 3000.jospoortvliet - Wednesday, November 22, 2017 - link
You need competition, plain as that. There isn't much of that in eu and us right now... hence India is a decade ahead in gb per dollar.FullmetalTitan - Thursday, November 23, 2017 - link
5G infrastructure definitely won't incentivize telecoms to increase caps immediately, but think back to pre-smart phone days for internet access on mobile devices versus the typical data plan size now (and effective cost in $/MB or $/GB). I agree that short term they will use 5G as a means to widen their margins, but long term the pricing window should shift towards larger caps.The secondary implications of effective mesh networks for diagnostic use and applications like vehicle-to-vehicle communication are probably a more important effect of 5G development.
ddrіver - Monday, November 20, 2017 - link
Data caps exist because companies are greedy. Period. When companies aren't greedy they upgrade their network.Reflex - Monday, November 20, 2017 - link
I tend to agree with this sentiment when it comes to wired networks. Upgrading them is a straight up investment that pays off over time. Wireless is a different case, you are restricted by available spectrum and the spec in use. LTE does not permit you to put as many towers as you want into an area, they must e a minimum distance from each other, and the performance of those towers is restricted by the specification and the frequencies a given carrier owns.Usage caps make far more sense in the wireless space than they do in the wired arena.
jtcarver - Monday, November 20, 2017 - link
Wired networks suffer from congestion just like wireless networks, even fiber optic cable has limits on the amount of data that can flow through the pipe at any given time, hence the data caps on wired networks that is cropping up these days, Cox Cable implements a 1tb data cap on most plans these days, DSL hasn't yet that I have seen, but their speed on regular DSL is so slow that you would be hard pressed to reach 1tb in a month.WizardMerlin - Thursday, November 23, 2017 - link
They do, but nothing like the same. You can put Terabits/s down a pair of fibres - you can also add more fibres as close as you like without causing a problem, you can't do that with wireless, the spectrum is the spectrum, you can't add more 5Ghz, the 5Ghz is the 5Ghz you've got and that's it - no matter how much you money you might have for upgrades; if it's congested it's congested.ads295 - Tuesday, November 21, 2017 - link
While I believe from the bottom of my heart that capitalism as a system begets greed and I do agree with you, there might be more to the picture:What if cellular networks here can be monetized to a greater degree on account of a bigger user base, thus allowing higher revenue from (say) the same area? So in the US say you'd have a 100 people using a network and here you'd have say 400. Thus data limits are the only way to get the sort of revenue you want (profit margins could be anything) in a low user base density.
Currently I pay INR 400 for 70 days of LTE, allowing 1GB per day of full speed data (and unlimited 128Kbps until midnight).
iwod - Monday, November 20, 2017 - link
Are you sure enough people are using LTE and doing heavy load on it? No Carrier has yet solved the user density problem in Major Cities areas, especially in high density region like South East Asia.Massive MIMO are now being deployed, and for TD-LTE Network that part of the equation is pretty much solved, as it brings 10x capacity improvement with No user equipment upgrade.
Massive MIMO on FD-LTE, where 70% of the world uses, is pretty much useless at this stage.
ads295 - Tuesday, November 21, 2017 - link
I've used a very popular LTE carrier in a very very high density place (a suburban train station that's one of the busiest in the world). And I can't even load a webpage there. But this case is actually more extreme than you'd see in more developed countries, right? I've moved back to a town (relatively speaking) and now speeds are all rightyannigr2 - Monday, November 20, 2017 - link
You are right, but. Better see companies create a network of 5G tomorrow, than expand a network of 3G or 4Gtoday . Because today is just a numbers game. In 10-15 years, consumers will probably be happy seeing at least a 5G in their connection status.Jon Tseng - Monday, November 20, 2017 - link
1) Hello Apple (let's face it, who else uses a discrete baseband nowadays, at least in a smartphone?).2) Does the 7660 support CDMA? Can't see any mention of this in the release (although it says the 8060 supports it). Was under the impression Intel were bringing in CDMA to displace Qualcomm but maybe it isn't ready yet?
Trixanity - Monday, November 20, 2017 - link
The current flagship modem 7560 supports CDMA so I don't see why its successor won't.iwod - Wednesday, November 22, 2017 - link
7560 Support CDMA, but given latest report show Apple will still be using Qualcomm for CDMA in their next generation of iPhone, this means CDMA support on Intel's first CDMA likely isn't good enough yet.So 7660 may be the Baseband that will completely displace Qualcomm in iPhone 2019. Given how Intel put out this announcement rather early I expect the design is pretty impressive.
melgross - Monday, November 20, 2017 - link
It’s nice to read about all of this while people around the world are stuck at about 20Mb/s,now, with a very few able to get to 50. Unless you’re in a testing area, and right near the tower, much higher than that is a bust.Nevertheless, Qualcomm is in trouble.
PeachNCream - Monday, November 20, 2017 - link
20Mb/s?! My wired internets at home are only 10 (thanks bunches DSL)...I wish I could get anything close to 20 on any connection and it's not like I live in the middle of nowhere either. *pouts*iwod - Monday, November 20, 2017 - link
Lots of Questions unanswered.Intel only has one customer for its Baseband and that is Apple. So we know for sure the XMM75 series will be in next iPhone, and presumably the XMM76 Series in 2019 iPhone.
Last time we knew the XMM75 will be Intel's first in house fabbed Baseband, and based on 14nm. So are we going to assume XMM76 being on 10nm?
Another Interesting point is this is the first time Intel has a pre announced Baseband that is better and faster then everyone else has on the Roadmap, that is including Qualcomm and Huawei. Although we are hitting diminishing of returns in terms of real benefits.
witeken - Monday, November 20, 2017 - link
XMM 7660 is 10nm, according to Ashraf Eassa.iwod - Wednesday, November 22, 2017 - link
I wouldn't bet on his word, he has been wrong about Intel a lot more then KGI on Apple. ( And sometimes his word dont make any sense )DigitalFreak - Monday, November 20, 2017 - link
From what I've read, their current baseband used in the iPhone sucks at poorer signal levels. Hopefully these new ones won't have that issue.iwod - Wednesday, November 22, 2017 - link
That is the first Intel Baseband on iPhone. No one has tested the newer one yet. Although I doubt it will be "better" then Qualcomm, but likely improve to a point where it makes little difference in real world.jtcarver - Monday, November 20, 2017 - link
Qualcomm had had 5g test networks in place for several months now, everything I read in this article that will be in Intel's 5g modem is already in Qualcomms modem, and there is a typo in the article where they mention a 58ghz band (there is no such thing, as that would be in x-ray territory). Since Qualcomm will be shipping their modem in 2018 and test units are already out to phone manufacturers it would seem that Intel is a little behind. If Intel really wanted to get their feet into the SoC and 5g modem business they would have taken more of a role in developing the 5g standards rather than letting Qualcomm be the major player and getting all the standards based patentsiwod - Wednesday, November 22, 2017 - link
Well Intel is partnering with everybody ( apart from Qualcomm ) to develop the 5G. So the battle on Standard based patents is a lot more complicated, including lots of politics.There is no defined "5G" yet. So what everyone is talking about right now is mmWave. But before any of mmWave happens, carrier has to implement Massive MIMO and LAA first, as well as many other LTE Advance Pro, 4.9G or 3GPP Rel 14 / 15 before 5G happens.
As a matter of fact I think Carrier will likely brand their 4.9G Network as 5G, and then let 5G evolves from it. Much like what happen with 4G.
WizardMerlin - Thursday, November 23, 2017 - link
58Ghz is nowhere near X-ray territory, up to 300Ghz is still classed as microwaves. 802.11ad is a 60Ghz wireless standard.Morawka - Monday, November 20, 2017 - link
All I can say is FINALLY! Qualcomm now has some real competition. Intel is first out of the gate with 5G so they will get their patents filed early. Now Intel needs to work on integrating and licensing out these modems for use in SOC's instead of just offering them as a standalone modem. Integration is important for mobile as precious PCB space is not taken up by the modem.jtcarver - Monday, November 20, 2017 - link
To little too late, Qualcomm already has test 5g networks in place, has had for several months now and their 5g modem will be incorporated into their next SoC series (not 2 years from now), Qualcomm was the major player in implementing the 5g standards and already patented their standards based 5g technology. Everything that I read in this article is already in Qualcomms 5g modem that will be shipping in 2018. Intel is a newcomer in the mobile communications business and their only customer that really matters is apple, Samsung uses Qualcomm or their own SoC and modems, Huawei uses either their own SoC or Qualcomm, and are designing their own 5g modem so that leaves Intel's market share pretty thinSydneyBlue120d - Tuesday, November 21, 2017 - link
So are you sain' that direct X20 successor will be a full multimode 5G-4G-3G-2G modem?!?!!?yeeeeman - Saturday, November 25, 2017 - link
I really don't get why they bother so much improving maximum speed supported by modems each year when the actual network doesn't even support a quarter of the 1Gbps at best.The industry today has taken the obsessive route of improving things at a very fast and unnecessary rate, just for the sake of better spec (read higher numbers) on a new device so that people will be motivated to buy it. Stupid really...
SydneyBlue120d - Monday, November 27, 2017 - link
Here you find the answer:https://semiaccurate.com/2017/11/14/qualcomm-shows...
Estyle - Saturday, July 7, 2018 - link
Any 5G devices on this chipset now?