When the hell is 14.4 Mbps coming to landline broadband, let alone wireless broadband? Actually, come to think of it, 7.2 Mbps is faster than my cable connection at home. Something is seriously wrong in the ISP business...
I currently have a consistently stable 16Megabit connection via Charter Cable. It used to be a 10Meg line, but they changed their packages and every 10Meg user was moved to the 16Meg package. They also offer a 25Meg line.
High speed with no coverage is the likely result. AT&T coverage is terrible. Fast terrible is still terrible. Dropped calls, poor consumer experience will continue in my estimation. Waiting for an alternate iPhone carrier. Someone please rescue us!
I get plenty of AT&T signal....just rarely get throughput in spite of how many bars I have. I can't count how many times a day I have to tap OK to "Unable to reach server" on my iPhone. Can AT&T please fix 3G before they start spending all this $ on upgrading to the next great thing?
Possibly it is not the 3G signal but the back end internet connection that is the problem. Two different things. It's like having a wireless router in your house. You may have a great connection to the wireless router but if that has a poor internet connection you'll won't see high speeds.
I know people with broadband cable that is faster than the theoretical speeds of 802.11g (54Mbps). It's at the point where consumer router processing abilities are the bottleneck. I hope Apple updates the AirPort Extreme Base Station to deal with this growing issue.
Doesn't the Airport Extreme already offer simultaneous dual-band with 802.11n running on the 5GHz band? I'm running a Time Machine as the primary wireless router with an Extreme as an extender and two AirPort Express units in the den and master bedroom for music. LAN speed doesn't seem ever to be a problem in our house, just periodic slowdowns in Comcast broadband. Four Macs are doing hourly Time Capsule backups that can be up to 500MB or more on our LAN. The only processing and download slowdowns I see involve the slower Macs and MacBooks when they are backing up to Time Capsule at the time. I suspect ISPs are at the root of most consumer complaints about slow download performance, not LAN speeds.
BTW, no complaints in the Chicago area when our iPads are connected 3G to AT&T Wireless. In fact, that's been the case anywhere they've shown five-bar 3G signal strengths - Ft. Lauderdale, St. Augustine, Savannah, Richmond, Washington metro. Fast downloads and uploads almost everywhere there's solid 3G. The AT&T national data coverage map was right in line with our experience - long no-3G stretches through rural sections of I-95 in Georgia and the Carolinas, and I-80 across Pennsylvania. 3G data performance was not quite as good with our 18-month-old iPhone 3G models as with the iPads. The newer, faster iPad hardware appears to have a lot to do with it. It'll be interesting to see the early reports on the upcoming 4G iPhone wireless performance.
It is interesting that dropped calls and cell tower capacity figure so much. I have been fortunate enough to use many GSM/UMTS networks around the world and have always noted that the towers work on the principle of 3. 3 sides with 3 panes/Antennae, or 3 panes on a pole.
I'm not a cell tower engineer (i.e. a radio network planner), but Do have some background on the matter. Elsewhere you might be seeing three sectors with three different frequencies in use. WCDMA uses 5MHz bands (actually 3.84 Mhz + guard) and one way to add capacity to an area is to add additional 5MHz carriers to an existing 3 sector site. One of the issues in the US (why you see only six) could be the limited number of 5MHz carriers.
The other aspect of course is the physical coverage thing that may affect your reception and dropped calls. The third is the iPhone's known bad WCDMA implementation.
About HSPA+: 14Mbps is NOT HSPA+. HSPA goes up to 14Mbps on its own (14Mbps DL with 16QAM modulation). HSPA+ on the other hand allows bitrates of 21, 28 and 42 Mbps with 64 QAM, MIMO and Dual Cell combinations. This is what is being currently deployed worldwide. Eventually HSPA+ will achieve bitrates of 168Mbps (theoretical) and talk is of reaching 300Mpbs eventually.
The bitrates that you see however are very greatly affected by the radio conditions and for example HSPA+ @ 21mbps rate is in practise no better than HSPA with 14.4 Mbps because it is very rare to get good enough radio conditions to achieve the 21mbps with the higher order modulation of 64QAM. The way networks are planned in general, you should not see more than 4-5 Mbps on average unless you use MIMO or Dual Cell. iPhone is not capable of this (yet). Neither apparently is AT&T. For the phone, a Firmware update is not going to do it, for the network it might if the equipment is from the right vendor and the HW is new enough.
Also if your network is only capable of high speed in the downlink (i.e. HSDPA), then you will never see a uplink speed of higher than 384kbps. You'd need HSUPA for that. HSDPA and HSUPA together are known as HSPA (High Speed Packet Access).
I don't get it. AT&T will have a theoretical 10mbps+ network , but the service will still suck well into the next few years. Good bl**dy luck with a decent signal, decent connection and decent speeds. Telcos are the next big industry mammoths that need a kick up the butt after Big Oil.
High speed with no coverage is the likely result. AT&T coverage is terrible. Fast terrible is still terrible. Dropped calls, poor consumer experience will continue in my estimation. Waiting for an alternate iPhone carrier. Someone please rescue us!
It's all about marketing. They won't be able to claim "The Fastest 3G" in their ads if they don't do this. They don't claim the most reliable or the most coverage just the fastest.
Bingo x 2
It's so obvious this is primarily a Marketing/PR stunt with marginal, slow-to-realise engineering achievements and user experience benefits.
Still can't really get internet access on my iphone at a Phillies game. Have to switch my 3GS to Edge to get a connection. Yes I realize that 11,000 iphones in one spot might be a bit of challenge, but it's just that, a challenge figure it out ATT before we hit you with a Tazer and then puke on you!
That's the bitch of it. ATT can't even get 1,000 + iphones to work in one small concentrated area. Yet, they are trying to up speeds. They need to work on reliability.
Doesn't the Airport Extreme already offer simultaneous dual-band with 802.11n running on the 5GHz band? I'm running a Time Machine as the primary wireless router with an Extreme as an extender and two AirPort Express units in the den and master bedroom for music. LAN speed doesn't seem ever to be a problem in our house, just periodic slowdowns in Comcast broadband. Four Macs are doing hourly Time Capsule backups that can be up to 500MB or more on our LAN. The only processing and download slowdowns I see involve the slower Macs and MacBooks when they are backing up to Time Capsule at the time. I suspect ISPs are at the root of most consumer complaints about slow download performance, not LAN speeds.
I'm talking about what the WiFi on the router can potentially do, I'm talking about what the router can actually do. The processing abilities of these consumer routers aren't powerful enough to deal with the speed of many broadband internet services or the growing number of computers on networks.
Scenario 1: Take a 100Mbps cable connection. Connect the modem to your machine using a 1000BASE-T Ethernet port, bypassing the router. Test the throughput.
Scenario 2: Take a 100Mbps cable connection. Connect the modem to your router, then test the throughput via the 1000BASE-T switch to your machine. Test the throughput.
Scenario 3: Take a 100Mbps cable connection. Connect the modem to your router, then test the throughput via 802.11n to your machine. Test the throughput.
Scenario 4: Take a 100Mbps cable connection. Connect the modem to your router, then test the throughput via 802.11n to your machine. Also, connect more machines via the router and have each of them transfer copious amounts of data to each other over the LAN. Test the throughput.
In each scenario the processing the processing overhead is increased, thus slowing down the usable connection. Even in scenario one this can be seen as the AEBS isn't powerful enough to make good use of a 100Mbps connection. The switch being 1Gbps or 802.11n being 108Mbps is irrelevant to the discussion. I can put racing tires on my car rated for x-speed, but that means nothing if my engine is crap.
Even in scenario one this can be seen as the AEBS isn't powerful enough to make good use of a 100Mbps connection. The switch being 1Gbps or 802.11n being 108Mbps is irrelevant to the discussion. I can put racing tires on my car rated for x-speed, but that means nothing if my engine is crap.
Thanks for your response. In that context I wonder if AI can turn up any info on upcoming Apple router product improvements.
Thanks for your response. In that context I wonder if AI can turn up any info on upcoming Apple router product improvements.
It's my hope (and of others on this site) that Apple will use iPhone OS in an upcoming A4-based AEBS, Time Capsule, Apple Home Server and AppleTV. Only the latter two would clearly benefit from iPhone OS. The first two I'm less certain about.
Of course, I don't mean Cocoa Touch as that is pointless for these networked devices, but the foundation of iPhone OS. Everything except the AppleTV could get away with not including the GPU, but if Apple is buying in bulk it might be cheaper just to leave it in their.
The AppleTV currently uses a special 1GHz Pentium M "Crofton" CPU with 256MB Memory running pretty much standard Mac OS X with much tweaking. Sure, they removed some things and added others, but it's far from being an ideal OS for the HW like in their other OS X-based devices.
The Time Capsule uses a Marvell 88F5BF01 C500 500Mhz CPU with 128MB RAM and 16MB for the OS.
It amazes me how out of touch AI and the commentors can be.
The reality is that AT&T upgraded the equipment on all 3G towers to HSPA 7.2 last year. They did not enable the bandwith to handsets on all towers however, because they also needed to upgrade the backhaul connections which is taking more time.
The HSPA+ upgrade is a software upgrade, and not a huge capital investment. It will however give improvements other than just bandwidth, including improved reliability. They are going to run it at a max 14.4 not the 21 some other networks are offering to data cards, but it is still HSPA+. They will be spending 18-19 billion on infrastructure upgrades this year, and I would expect much of it will be in the backhaul which will allow them to "turn on" the 14 mb software. These upgrade should therefore have significant noticable benefits to all 3G users including those with iPhone 3G and 3GS handsets that don't support the 14.4 speeds. In fact the benefits are already showing up.
All the naysayers and Verizon fanboys need to look at the facts. For example, PC World's real devices in the wild tests this winter showed AT&T's real world download speeds doubled from their test last Spring. They were up in every city tested. In contrast, Verizon was DOWN in every city tested. Odly enough the reduction in Verizon's performance corresponds with the rollout of the Droid and a few other phones that are FINALLY encouraging some select Verizon customers to use data volumes like AT&T has had to deal with. In other words we don't need to wait for an iPhone on Verizon to know it will overload their network, we are seeing it already with a much much smaller install base of Android phones. Good smartphones kill networks. AT&T has been hit first and hit the hardest, but no one is immune.
Another interesting note is that AT&T's upload speeds on the HSPA 7.2 network tested faster than Sprint's 4G upload speeds. Their new HSPA+ speeds, down and up, are expected to be comparable to the real world speeds Verizon will get from their initial LTE deployment, except AT&T will be able to deliver the network and phones this year, whereas Verizon will have the network this year but no phones until sometime in 2011. Of course you still need a device that can actually handle and make use of real data rates in excess of 5 mbps, even if the chipsets and the network can theoretically handle it. Don't fall for the branding or the marketing, 4G is still a ways off from delivering real benefits to the consumers phone, and the better 3G technologies still have some life in them. Data cards are a slightly different story, with earlier availability and devices that can use much more bandwidth.
AT&T still has coverage issues, both in congested cities and in the "map" of 3G coverage at any speed. I am not sure how much of that will get addressed before the 700 mhz LTE rollout, but that definitely remains a valid factually based complaint about AT&T.
In short, AT&T has done a lot to improve their network already, and this HSPA+ news is good news for those who are still having issues, even if you are not planning on getting a new device. While Verizon has better rural coverage, their network appears to be less robust than AT&T's and the congestion issues are starting to show up now that they have a growing base of usable smartphones. The much hoped for Verizon iPhone will probably not make everyone as happy as they think, because if it gets released and sells by the millions, the network performance will suffer. AT&T may actually enjoy seeing that!
The HSPA+ upgrade is a software upgrade, and not a huge capital investment.
It will however give improvements other than just bandwidth, including improved reliability. They are going to run it at a max 14.4 not the 21 some other networks are offering to data cards, but it is still HSPA+.
If it's not offering 64 QAM, MIMO or Dual Carrier, it ain't HSPA+, it's just regular HSPA which allows up to 14 Mbps with 15 codes used simultaneously on one carrier using 16QAM modulation. Regular HSPA has been rolled out in many places starting with 3.6 Mbps, then later with SW upgrades to 7.2 and then 10.8 or straight to 14 depending on operator and country.
I know it sounds like semantics, but the difference (especially if you do MIMO) but once you get full HSPA+ you'll see a clear improvement even in relatively poor radio conditions assuming of course that you have the latest and greatest phone that supports it (they don't really exist yet).
If it's just a regular HSPA with 14 mbps "enabled", the changes aren't that big especially if your phone isn't capable of using more than 10 codes simultaneously. It will however increase the cell's total throughput by 50% so if it's congested due to multiple users, it will improve things.
My understanding is that the iPhone is a category 8 phone (UE), thus maxing out at 7.2 Mbps theoretical:
If it's not offering 64 QAM, MIMO or Dual Carrier, it ain't HSPA+, it's just regular HSPA which allows up to 14 Mbps with 15 codes used simultaneously on one carrier using 16QAM modulation.
It will be HSPA+ with 64QAM based on what I am reading. No mention of MIMO yet, but MIMO is not necessary to be HSPA+ and certainly not necessary to benefit existing smart phones. If they can get real world speeds of 6-12 mb down, they will be able to match or beat Verizon's early LTE offering and Sprint's 4G. T-Mobile may get a leg up on all three with the speed of their HSPA+ but their network has other issues that will keep them from being a threat to AT&T or Verizon for a long time to come. More importanly, if the iPhone 3G can consistently average over 2 mb down and they can get the 3GS average into the 4+ range they will be way ahead of where Verizon is with the Droid (around 1.1).
FWIW, I am getting 1.8-2.2 mb down on my iPhone 3G today in my office which is not next to an external wall. According to my Speedtest logs, my best 3G speeds were 1.4-1.6 mb in April-May 2009 and often I was geting 0.8 to 1.1. I saw a jump to the 2mb range in the fall of 2009 in my area. I was not necessarily that noticeable in normal usage, but it definitely showed up in the Speedtest results. I get a max of 4.5-5.5 over WIFI on the iPhone 3G. My wife's 3GS is generally faster on both 3G and WIFI, so as should be expected, the phone is as important as the net. I am up for a new iPhone next week, so I am definitely looking forward to the iPhone 4g/HD release to see what kind of real world speeds I can get with that.
What this really means is for phone users (not laptop cards) the fastest networks from the second half of 2010 through the first half of 2011 will likely be T-Mobile and AT&T followed closely by Sprint's 4G. Verizon will lag far behind the pack until they can get LTE handsets and decent LTE coverage. If you were wondering why Verizon is rushing to get onto LTE before the hardware is even ready, here is your answer. I am sure 9-12 months in 4th place will not sit well with their marketing guys. If Verizon continues to have increasing speed and reliability issues due to congestion from the new Android phones, coupled with their slower 3G speeds, AT&T marketing will have a fun time getting revenge for those "There's a map for that" adds. A few jabs at the crazy high ETF for the Droid that downloads at half the speed of a 2008 iPhone would be quite entertaining. Could you imagine if AT&T can trully make their network that much better and if Verizon's non-LTE network does get worse, how that would impact a potential Verizon iPhone release in 2011 or 2012? It will be very interesting to see how the non-techy consumer sees things over the next 18-24 months.
When the hell is 14.4 Mbps coming to landline broadband, let alone wireless broadband? Actually, come to think of it, 7.2 Mbps is faster than my cable connection at home. Something is seriously wrong in the ISP business...
We have 10Mbps & 20Mbps cable here where I live, you just got a crappy cable company.
More bandwidth is useless if they still cap at 5GB download a month! And when are we gonna get tethering?!
When the hell is 14.4 Mbps coming to landline broadband, let alone wireless broadband? Actually, come to think of it, 7.2 Mbps is faster than my cable connection at home. Something is seriously wrong in the ISP business...
It will be HSPA+ with 64QAM based on what I am reading. No mention of MIMO yet, but MIMO is not necessary to be HSPA+ and certainly not necessary to benefit existing smart phones.
Correct on the MIMO not required to be HSPA+. The AI article (or engadget) didn't tell if 64QAM or MIMO was supported. Still sounds odd that they'd only go 14,4 as with 64 QAM you could go 21. I know about the backhaul, but who in their right minds would increase their backhaul to handle 14.4 from 7.2, but not put more knowing full well that in a few short years time, we are talking 40-130Mbps speeds?
With 64 QAM alone, the real benefit in real networks is negligible (5-15% improvement in speed). It also will require new terminals (phones) as the chipsets in current phones don't do that.
The real world Channel Quality Indicator values (CQI) that you'll see will force the HSPA+ with only 64QAM modulation to perform roughly the same as it would if you were running 16QAM. Have done and seen enough tests to be a believer in that. Once we get MIMO in the network and terminals and perhaps Dual Carrier, then you'll see real improvements.
Of course if the iPhone HD or whatever it will be called can use 15 codes instead of the current 10, you'll see a performance increase in cells where you are alone.
Still sounds odd that they'd only go 14,4 as with 64 QAM you could go 21.
Yes, and for marketing alone it would seem to make sense to claim "up to 21" like T-Mobile and the Canadians even though you will really only get 14 if you are not right next to a tower. That is 14 if you are lucky! There may be some coverage benefits to not going to 21, I have no clue. Either way, if the actual real world speeds are over 7.2 that will be a heck of an achievement based on what we were getting 12 months ago, and for that matter what we are getting now, knowing that there will be even more iPhones and other smartphones on their net by early summer. More important than bandwidth will be latency reductions and reliability improvements. AT&T has gotten much better on both counts in the last year, even when I venture into the edge area in the NY finger lakes region. Sure would love to see more 3G reach, but I am not sure how many subscribers they have amongst the amish and corn farmers.
Comments
When the hell is 14.4 Mbps coming to landline broadband, let alone wireless broadband? Actually, come to think of it, 7.2 Mbps is faster than my cable connection at home. Something is seriously wrong in the ISP business...
I currently have a consistently stable 16Megabit connection via Charter Cable. It used to be a 10Meg line, but they changed their packages and every 10Meg user was moved to the 16Meg package. They also offer a 25Meg line.
I get plenty of AT&T signal....just rarely get throughput in spite of how many bars I have. I can't count how many times a day I have to tap OK to "Unable to reach server" on my iPhone. Can AT&T please fix 3G before they start spending all this $ on upgrading to the next great thing?
Possibly it is not the 3G signal but the back end internet connection that is the problem. Two different things. It's like having a wireless router in your house. You may have a great connection to the wireless router but if that has a poor internet connection you'll won't see high speeds.
I know people with broadband cable that is faster than the theoretical speeds of 802.11g (54Mbps). It's at the point where consumer router processing abilities are the bottleneck. I hope Apple updates the AirPort Extreme Base Station to deal with this growing issue.
Doesn't the Airport Extreme already offer simultaneous dual-band with 802.11n running on the 5GHz band? I'm running a Time Machine as the primary wireless router with an Extreme as an extender and two AirPort Express units in the den and master bedroom for music. LAN speed doesn't seem ever to be a problem in our house, just periodic slowdowns in Comcast broadband. Four Macs are doing hourly Time Capsule backups that can be up to 500MB or more on our LAN. The only processing and download slowdowns I see involve the slower Macs and MacBooks when they are backing up to Time Capsule at the time. I suspect ISPs are at the root of most consumer complaints about slow download performance, not LAN speeds.
BTW, no complaints in the Chicago area when our iPads are connected 3G to AT&T Wireless. In fact, that's been the case anywhere they've shown five-bar 3G signal strengths - Ft. Lauderdale, St. Augustine, Savannah, Richmond, Washington metro. Fast downloads and uploads almost everywhere there's solid 3G. The AT&T national data coverage map was right in line with our experience - long no-3G stretches through rural sections of I-95 in Georgia and the Carolinas, and I-80 across Pennsylvania. 3G data performance was not quite as good with our 18-month-old iPhone 3G models as with the iPads. The newer, faster iPad hardware appears to have a lot to do with it. It'll be interesting to see the early reports on the upcoming 4G iPhone wireless performance.
It is interesting that dropped calls and cell tower capacity figure so much. I have been fortunate enough to use many GSM/UMTS networks around the world and have always noted that the towers work on the principle of 3. 3 sides with 3 panes/Antennae, or 3 panes on a pole.
I'm not a cell tower engineer (i.e. a radio network planner), but Do have some background on the matter. Elsewhere you might be seeing three sectors with three different frequencies in use. WCDMA uses 5MHz bands (actually 3.84 Mhz + guard) and one way to add capacity to an area is to add additional 5MHz carriers to an existing 3 sector site. One of the issues in the US (why you see only six) could be the limited number of 5MHz carriers.
The other aspect of course is the physical coverage thing that may affect your reception and dropped calls. The third is the iPhone's known bad WCDMA implementation.
About HSPA+: 14Mbps is NOT HSPA+. HSPA goes up to 14Mbps on its own (14Mbps DL with 16QAM modulation). HSPA+ on the other hand allows bitrates of 21, 28 and 42 Mbps with 64 QAM, MIMO and Dual Cell combinations. This is what is being currently deployed worldwide. Eventually HSPA+ will achieve bitrates of 168Mbps (theoretical) and talk is of reaching 300Mpbs eventually.
The bitrates that you see however are very greatly affected by the radio conditions and for example HSPA+ @ 21mbps rate is in practise no better than HSPA with 14.4 Mbps because it is very rare to get good enough radio conditions to achieve the 21mbps with the higher order modulation of 64QAM. The way networks are planned in general, you should not see more than 4-5 Mbps on average unless you use MIMO or Dual Cell. iPhone is not capable of this (yet). Neither apparently is AT&T. For the phone, a Firmware update is not going to do it, for the network it might if the equipment is from the right vendor and the HW is new enough.
Also if your network is only capable of high speed in the downlink (i.e. HSDPA), then you will never see a uplink speed of higher than 384kbps. You'd need HSUPA for that. HSDPA and HSUPA together are known as HSPA (High Speed Packet Access).
Regs, Jarkko
High Speed Garbage
High speed with no coverage is the likely result. AT&T coverage is terrible. Fast terrible is still terrible. Dropped calls, poor consumer experience will continue in my estimation. Waiting for an alternate iPhone carrier. Someone please rescue us!
Bingo.
It's all about marketing. They won't be able to claim "The Fastest 3G" in their ads if they don't do this. They don't claim the most reliable or the most coverage just the fastest.
Bingo x 2
It's so obvious this is primarily a Marketing/PR stunt with marginal, slow-to-realise engineering achievements and user experience benefits.
Still can't really get internet access on my iphone at a Phillies game. Have to switch my 3GS to Edge to get a connection. Yes I realize that 11,000 iphones in one spot might be a bit of challenge, but it's just that, a challenge figure it out ATT before we hit you with a Tazer and then puke on you!
That's the bitch of it. ATT can't even get 1,000 + iphones to work in one small concentrated area. Yet, they are trying to up speeds. They need to work on reliability.
Doesn't the Airport Extreme already offer simultaneous dual-band with 802.11n running on the 5GHz band? I'm running a Time Machine as the primary wireless router with an Extreme as an extender and two AirPort Express units in the den and master bedroom for music. LAN speed doesn't seem ever to be a problem in our house, just periodic slowdowns in Comcast broadband. Four Macs are doing hourly Time Capsule backups that can be up to 500MB or more on our LAN. The only processing and download slowdowns I see involve the slower Macs and MacBooks when they are backing up to Time Capsule at the time. I suspect ISPs are at the root of most consumer complaints about slow download performance, not LAN speeds.
I'm talking about what the WiFi on the router can potentially do, I'm talking about what the router can actually do. The processing abilities of these consumer routers aren't powerful enough to deal with the speed of many broadband internet services or the growing number of computers on networks.In each scenario the processing the processing overhead is increased, thus slowing down the usable connection. Even in scenario one this can be seen as the AEBS isn't powerful enough to make good use of a 100Mbps connection. The switch being 1Gbps or 802.11n being 108Mbps is irrelevant to the discussion. I can put racing tires on my car rated for x-speed, but that means nothing if my engine is crap.
Those are up-to speeds. Isn't throttling commonplace in Canada? I'd also exceed the 75GB cap in a week, or less.
FIOS is the way to go, where it is available-
You must be working with the government.
Verizon's FiOS bundles, i.e., Tier 3 (35MBs Down and Up) is regularly $110/month.
And what are you doing to use so much data? Are you uploading much?
Even in scenario one this can be seen as the AEBS isn't powerful enough to make good use of a 100Mbps connection. The switch being 1Gbps or 802.11n being 108Mbps is irrelevant to the discussion. I can put racing tires on my car rated for x-speed, but that means nothing if my engine is crap.
Thanks for your response. In that context I wonder if AI can turn up any info on upcoming Apple router product improvements.
Thanks for your response. In that context I wonder if AI can turn up any info on upcoming Apple router product improvements.
It's my hope (and of others on this site) that Apple will use iPhone OS in an upcoming A4-based AEBS, Time Capsule, Apple Home Server and AppleTV. Only the latter two would clearly benefit from iPhone OS. The first two I'm less certain about.
Of course, I don't mean Cocoa Touch as that is pointless for these networked devices, but the foundation of iPhone OS. Everything except the AppleTV could get away with not including the GPU, but if Apple is buying in bulk it might be cheaper just to leave it in their.
The AppleTV currently uses a special 1GHz Pentium M "Crofton" CPU with 256MB Memory running pretty much standard Mac OS X with much tweaking. Sure, they removed some things and added others, but it's far from being an ideal OS for the HW like in their other OS X-based devices.
The Time Capsule uses a Marvell 88F5BF01 C500 500Mhz CPU with 128MB RAM and 16MB for the OS.
The reality is that AT&T upgraded the equipment on all 3G towers to HSPA 7.2 last year. They did not enable the bandwith to handsets on all towers however, because they also needed to upgrade the backhaul connections which is taking more time.
The HSPA+ upgrade is a software upgrade, and not a huge capital investment. It will however give improvements other than just bandwidth, including improved reliability. They are going to run it at a max 14.4 not the 21 some other networks are offering to data cards, but it is still HSPA+. They will be spending 18-19 billion on infrastructure upgrades this year, and I would expect much of it will be in the backhaul which will allow them to "turn on" the 14 mb software. These upgrade should therefore have significant noticable benefits to all 3G users including those with iPhone 3G and 3GS handsets that don't support the 14.4 speeds. In fact the benefits are already showing up.
All the naysayers and Verizon fanboys need to look at the facts. For example, PC World's real devices in the wild tests this winter showed AT&T's real world download speeds doubled from their test last Spring. They were up in every city tested. In contrast, Verizon was DOWN in every city tested. Odly enough the reduction in Verizon's performance corresponds with the rollout of the Droid and a few other phones that are FINALLY encouraging some select Verizon customers to use data volumes like AT&T has had to deal with. In other words we don't need to wait for an iPhone on Verizon to know it will overload their network, we are seeing it already with a much much smaller install base of Android phones. Good smartphones kill networks. AT&T has been hit first and hit the hardest, but no one is immune.
Another interesting note is that AT&T's upload speeds on the HSPA 7.2 network tested faster than Sprint's 4G upload speeds. Their new HSPA+ speeds, down and up, are expected to be comparable to the real world speeds Verizon will get from their initial LTE deployment, except AT&T will be able to deliver the network and phones this year, whereas Verizon will have the network this year but no phones until sometime in 2011. Of course you still need a device that can actually handle and make use of real data rates in excess of 5 mbps, even if the chipsets and the network can theoretically handle it. Don't fall for the branding or the marketing, 4G is still a ways off from delivering real benefits to the consumers phone, and the better 3G technologies still have some life in them. Data cards are a slightly different story, with earlier availability and devices that can use much more bandwidth.
AT&T still has coverage issues, both in congested cities and in the "map" of 3G coverage at any speed. I am not sure how much of that will get addressed before the 700 mhz LTE rollout, but that definitely remains a valid factually based complaint about AT&T.
In short, AT&T has done a lot to improve their network already, and this HSPA+ news is good news for those who are still having issues, even if you are not planning on getting a new device. While Verizon has better rural coverage, their network appears to be less robust than AT&T's and the congestion issues are starting to show up now that they have a growing base of usable smartphones. The much hoped for Verizon iPhone will probably not make everyone as happy as they think, because if it gets released and sells by the millions, the network performance will suffer. AT&T may actually enjoy seeing that!
The HSPA+ upgrade is a software upgrade, and not a huge capital investment.
It will however give improvements other than just bandwidth, including improved reliability. They are going to run it at a max 14.4 not the 21 some other networks are offering to data cards, but it is still HSPA+.
If it's not offering 64 QAM, MIMO or Dual Carrier, it ain't HSPA+, it's just regular HSPA which allows up to 14 Mbps with 15 codes used simultaneously on one carrier using 16QAM modulation. Regular HSPA has been rolled out in many places starting with 3.6 Mbps, then later with SW upgrades to 7.2 and then 10.8 or straight to 14 depending on operator and country.
I know it sounds like semantics, but the difference (especially if you do MIMO) but once you get full HSPA+ you'll see a clear improvement even in relatively poor radio conditions assuming of course that you have the latest and greatest phone that supports it (they don't really exist yet).
If it's just a regular HSPA with 14 mbps "enabled", the changes aren't that big especially if your phone isn't capable of using more than 10 codes simultaneously. It will however increase the cell's total throughput by 50% so if it's congested due to multiple users, it will improve things.
My understanding is that the iPhone is a category 8 phone (UE), thus maxing out at 7.2 Mbps theoretical:
http://www.radio-electronics.com/inf...data-rates.php
So 4-5Mbps on an iPhone with speedtest means that your radio network in your location is in very good condition.
Regs, Jarkko
If it's not offering 64 QAM, MIMO or Dual Carrier, it ain't HSPA+, it's just regular HSPA which allows up to 14 Mbps with 15 codes used simultaneously on one carrier using 16QAM modulation.
It will be HSPA+ with 64QAM based on what I am reading. No mention of MIMO yet, but MIMO is not necessary to be HSPA+ and certainly not necessary to benefit existing smart phones. If they can get real world speeds of 6-12 mb down, they will be able to match or beat Verizon's early LTE offering and Sprint's 4G. T-Mobile may get a leg up on all three with the speed of their HSPA+ but their network has other issues that will keep them from being a threat to AT&T or Verizon for a long time to come. More importanly, if the iPhone 3G can consistently average over 2 mb down and they can get the 3GS average into the 4+ range they will be way ahead of where Verizon is with the Droid (around 1.1).
FWIW, I am getting 1.8-2.2 mb down on my iPhone 3G today in my office which is not next to an external wall. According to my Speedtest logs, my best 3G speeds were 1.4-1.6 mb in April-May 2009 and often I was geting 0.8 to 1.1. I saw a jump to the 2mb range in the fall of 2009 in my area. I was not necessarily that noticeable in normal usage, but it definitely showed up in the Speedtest results. I get a max of 4.5-5.5 over WIFI on the iPhone 3G. My wife's 3GS is generally faster on both 3G and WIFI, so as should be expected, the phone is as important as the net. I am up for a new iPhone next week, so I am definitely looking forward to the iPhone 4g/HD release to see what kind of real world speeds I can get with that.
What this really means is for phone users (not laptop cards) the fastest networks from the second half of 2010 through the first half of 2011 will likely be T-Mobile and AT&T followed closely by Sprint's 4G. Verizon will lag far behind the pack until they can get LTE handsets and decent LTE coverage. If you were wondering why Verizon is rushing to get onto LTE before the hardware is even ready, here is your answer. I am sure 9-12 months in 4th place will not sit well with their marketing guys. If Verizon continues to have increasing speed and reliability issues due to congestion from the new Android phones, coupled with their slower 3G speeds, AT&T marketing will have a fun time getting revenge for those "There's a map for that" adds. A few jabs at the crazy high ETF for the Droid that downloads at half the speed of a 2008 iPhone would be quite entertaining. Could you imagine if AT&T can trully make their network that much better and if Verizon's non-LTE network does get worse, how that would impact a potential Verizon iPhone release in 2011 or 2012? It will be very interesting to see how the non-techy consumer sees things over the next 18-24 months.
When the hell is 14.4 Mbps coming to landline broadband, let alone wireless broadband? Actually, come to think of it, 7.2 Mbps is faster than my cable connection at home. Something is seriously wrong in the ISP business...
We have 10Mbps & 20Mbps cable here where I live, you just got a crappy cable company.
More bandwidth is useless if they still cap at 5GB download a month! And when are we gonna get tethering?!
Please no comments from jailbreakers.
When the hell is 14.4 Mbps coming to landline broadband, let alone wireless broadband? Actually, come to think of it, 7.2 Mbps is faster than my cable connection at home. Something is seriously wrong in the ISP business...
Where do you live?
It will be HSPA+ with 64QAM based on what I am reading. No mention of MIMO yet, but MIMO is not necessary to be HSPA+ and certainly not necessary to benefit existing smart phones.
Correct on the MIMO not required to be HSPA+. The AI article (or engadget) didn't tell if 64QAM or MIMO was supported. Still sounds odd that they'd only go 14,4 as with 64 QAM you could go 21. I know about the backhaul, but who in their right minds would increase their backhaul to handle 14.4 from 7.2, but not put more knowing full well that in a few short years time, we are talking 40-130Mbps speeds?
With 64 QAM alone, the real benefit in real networks is negligible (5-15% improvement in speed). It also will require new terminals (phones) as the chipsets in current phones don't do that.
The real world Channel Quality Indicator values (CQI) that you'll see will force the HSPA+ with only 64QAM modulation to perform roughly the same as it would if you were running 16QAM. Have done and seen enough tests to be a believer in that. Once we get MIMO in the network and terminals and perhaps Dual Carrier, then you'll see real improvements.
Of course if the iPhone HD or whatever it will be called can use 15 codes instead of the current 10, you'll see a performance increase in cells where you are alone.
Regs, Jarkko
Still sounds odd that they'd only go 14,4 as with 64 QAM you could go 21.
Yes, and for marketing alone it would seem to make sense to claim "up to 21" like T-Mobile and the Canadians even though you will really only get 14 if you are not right next to a tower. That is 14 if you are lucky! There may be some coverage benefits to not going to 21, I have no clue. Either way, if the actual real world speeds are over 7.2 that will be a heck of an achievement based on what we were getting 12 months ago, and for that matter what we are getting now, knowing that there will be even more iPhones and other smartphones on their net by early summer. More important than bandwidth will be latency reductions and reliability improvements. AT&T has gotten much better on both counts in the last year, even when I venture into the edge area in the NY finger lakes region. Sure would love to see more 3G reach, but I am not sure how many subscribers they have amongst the amish and corn farmers.