Receive new posts as email.
This site operates as an independent editorial operation. Advertising, sponsorships, and other non-editorial materials represent the opinions and messages of their respective origins, and not of the site operator. Part of the FM Tech advertising network.
Entire site and all contents except otherwise noted © Copyright 2001-2011 by Glenn Fleishman. Some images ©2006 Jupiterimages Corporation. All rights reserved. Please contact us for reprint rights. Linking is, of course, free and encouraged.
The folks at iFixIt found a dual-standard GSM/CDMA chip in the Verizon model of the iPhone 4: In Step 17, the teardown experts note that the Qualcomm MDM6600, which can support GSM standards up to HSPA+ (14.4 Mbps flavor) as well as Qualcomm's traditional CDMA voice and data standards up to EVDO Rev. A (deployed in the US by Verizon Wireless and Sprint Nextel) as well as EVDO Rev. B. There are an enormous number of shared characteristics between the GSM and CDMA standards, and Qualcomm represents a significant minority percentage of all the patents in a pool that's used for UMTS/HSPA.
Apple is spending more to put this dual-mode chip in, of course, than it would for a single-standard chip. But it makes a bloody lot of sense. By having a single chip that can be switched to GSM or CDMA, Apple can switch to a single line of manufacture to supply phones worldwide. They'll save the cost in a higher price for the chip by not having two separate products to make and track. I wouldn't be surprised if we see iPhone 4 models sold for the GSM market that are identical with new antenna locations to the so-called CDMA model.
Does this mean that Apple will offer a world phone for CDMA and GSM markets? Note that the Verizon version of the phone has no SIM slot nor built-in SIM card, so it can't be used on a GSM network in its current form even with a firmware update. Will an iPhone 5 be switchable? It's hard to tell. I imagine Verizon Wireless would prefer the CDMA lock in, but Verizon Wireless is minority-owned by Vodafone, a worldwide GSM provider, which would almost certainly like to sell a single model worldwide that could be easily switched to work in the US or in any of its non-US markets. There's a Droid that does that already.
So Verizon has an iPhone: Empires (of technology) have risen and fallen since Apple introduced its iPhone in 2007, and the question constantly on everyone's lips: When will Verizon get a model that works on its network? The answer: 3 February for existing customers and 10 February for new ones.
The CDMA-based iPhone has piles of tradeoffs, but these aren't necessarily worse than using a GSM iPhone on AT&T or other networks around the world.
The ViPhone doesn't have LTE. No phone has LTE. We're not going to see LTE-based phones with decent battery life and size for several months, and the early ones will be monsters of compromise. There is no agreed-upon voice standard for LTE networks yet, which means Verizon will make compromises in whatever voice option it picks (initially) before later upgrading to something more universally supported. I don't expect an LTE iPhone until 2012, because coverage and other tradeoffs won't make it desirable until then.
You can't talk and use 3G data at the same time. Verizon opted for EVDO (Evolution Data Only), which tells a story with its name. Voice is handled separately and can't be used simultaneously. Wi-Fi and voice can be used at the same time.
It's Verizon Wireless, for cripe's sake. Verizon has a history of offering less-than-forthcoming information about its service plans, and is in the middle of settling a dispute in which it denied for years charging people $1.99 and other fees for inadvertent usage of mobile data without a plan (when pressing a conveniently located button that's on every featurephone). People don't like any carrier, but Verizon didn't make itself any friends with this.
We don't know pricing plans yet. Verizon's 3G service plans aren't bad, but they aren't enormously better than AT&T's unless you use a ton of data each month.
Worldwide roaming isn't an option. Despite being 45-percent owned by GSM carrier Vodafone, the ViPhone won't work on most networks worldwide because it's CDMA only. It's odd that with Vodafone selling many millions of iPhones into other markets, Verizon Wireless, Vodafone, and Apple wouldn't have teamed up to offer both. There are chips, from Qualcomm notably, that allow GSM/CDMA switchover.
You won't be able to use your ViPhone in Canada or Mexico, notably, and you can't swap out a SIM in GSM markets for another carrier, something that carriers allow or make easy in some markets, or that you can jailbreak to allow. (Update: Canada and Mexico do have CDMA operators, and Verizon supports roaming. Coverage in Mexico is quite limited for voice and data; in Canada, there's broader availability, but 3G data isn't everywhere.)
AT&T has a path to faster service, with HSPA+ (21 Mbps) overlaying HSPA 7.2 (7.2 Mpbs) as the year goes on. Verizon is stuck at 3.1 Mbps EVDO Rev. A until it has a sufficient LTE footprint to jump its customers to that. (AT&T and Verizon will both require new hardware for faster networks, though. An iPhone 4 will not, to my knowledge, be firmware upgradable to HSPA+.)
If you live in an area with poor AT&T coverage and great Verizon coverage, which likely amounts to tens of millions of people's homes and workplaces around the US, then you are suddenly able to own an iPhone of your very own. AT&T's coverage can be sketchy in many parts of the country, notably exurbs and large parts of New England.
Verizon and Apple are offering a mobile hotspot feature (pricing not yet known), which lets the iPhone be used as a Wi-Fi hotspot, just like a MiFi, or Android 2.2 phone, or several other smartphones. This lets up to five devices connect. GSM networks and the GSM iPhone could support this feature, too. Apple has been keeping this in its back pocket, perhaps to save for Verizon. (I don't believe AT&T, unique among the carriers, has any phone with a mobile hotspot feature enabled
, nor does it offer a MiFi or similar router to the general public. Correction: AT&T reminded me via email that they had added a MiFi to their product line last November.)
Heavy data users will have an unlimited plan restored to them if Verizon offers its $30/mo unlimited data plan to new iPhone users. AT&T removed that plan for new subscribers in June, and has a 2 GB/mo plan for $25, with $10/GB overage fees (charged by the whole GB for any fraction).
Verizon will sell many millions of its CDMA iPhone to existing customers (upgrading from featurephones and sidegrading from other smartphones) and new customers (who have never owned a smartphone or are fleeing AT&T, cancellation fees be damned).
Verizon has the best and most solid 3G network in the US, proven time and again by independent third parties. AT&T has continually improved its network coverage and speed, and has great plans for 2011.
Can we have some competition now, please?
T-Mobile said today it would upgrade its HSPA+ network to 42 Mbps in 2011: Everyone keeps upping the ante. T-Mobile wants to persuade customers that it has the fastest network out there, and doubling its raw speed for HSPA+ from 21 Mbps to 42 Mbps is a good way to do it. T-Mobile invested in bringing high-rate backhaul to its 3G network (which it wants to call 4G; whatever), and this is how it pays off.
AT&T yesterday said it has HSPA+ everywhere, but its backhaul won't be fully in place at those sites even this year: only 2/3rds of HSPA+ sites will have capable bandwidth in 12 months' time, according to yesterday's AT&T press release.
T-Mobile has never given guidance on the percentage of its HSPA/HSPA+ network that has the necessary backhaul, although it's consistency talked about its intent to build that infrastructure as it developed its green-field 3G network a couple of years ago. (Update: Analyst Charles Golvin of Forrester Research wrote in to say that T-Mobile provided such information in a briefing yesterday: 70 percent of T-Mobile's sites have Ethernet. Ethernet doesn't imply a specific speed, but my understanding is that it's all being installed as gigabit Ethernet. Since a single site can have multiple HSPA+ channels in use, more than 100 Mbps is necessary.)
T-Mobile's upgrades are a defensive move against LTE in 700 MHz. T-Mobile has a relatively small spectrum portfolio in the US, and LTE networks in this country will launch with much wider channels, allowing greater capacity and higher speeds. The 700 MHz frequency range also allows better in-building and in-home penetration than T-Mobile's mobile data frequency allocations.
Put simply: T-Mobile is installing the most advanced current-generation, off-the-shelf equipment that it can to compete with next-generation networks that are barely off the ground (Verizon) or not at all, and will take until 2013 to have a complete footprint. T-Mobile has that edge. But the AT&T and Verizon LTE networks will have substantive advantages over HSPA+ because T-Mobile will need to install equipment at a much higher density than either competitor to achieve the same coverage and capacity.
T-Mobile also still only reaches 200m people with its current mobile data network. AT&T reaches about 50m more, and Verizon has over 95 percent national coverage with its older 3G technology.
AT&T offers specifics on its HSPA+ upgrade and LTE deployment: At CES today, AT&T released its timeline for rolling out 4G LTE mobile service, which launches in mid-2011. Verizon Wireless gets bragging rights with several markets lit up in late 2010. However, with few devices, and an odd pricing model for such a fast service, Verizon has very little lead over AT&T.
AT&T and Verizon will likely both complete national urban rollouts by 2013, the stated date by both firms now for that goal. Their various FCC licenses require either geographic or population based completion at four and eight year targets, which will drive LTE service into less-populated areas and small-to-medium-sized towns.
AT&T's current HSPA/HSPA+ network is also measurably faster than Verizon's, which cannot increase its 3G speed at all. AT&T, like T-Mobile, is taking advantage of baby steps with HSPA (to 7.2 Mbps) and HSPA+ (to 21 Mpbs) to have an interim advantage, as well as a better hybrid 3G/4G roaming experience. Verizon is stuck at EVDO Rev. A, about 3.1 Mbps downstream. (All those rates are raw, and Verizon's coverage area remains superior to AT&T's; HSPA+ doesn't offer an advantage if you can't actually pick up a signal.)
Of course, it wouldn't be a telecom announcement without having to pick apart some news. AT&T says that it has HSPA+ available to "virtually 100 percent of its mobile broadband network" but then notes it requires "Ethernet or fiber backhaul." It predicts 2/3rds of its HSPA+ footprint will have such "expanded backhaul" by the end of 2011.
Which means that at the beginning of 2011, substantially less than 2/3rds of AT&T HSPA+ network cannot deliver true HSPA+ speeds, being constrained by the backhaul. If it were more than half, you can believe AT&T would have stated that in the press release.
Carrier-grade operations are supposed to be carrier grade: In its enthusiasm to have LTE operating in multiple markets before year's end, Verizon Wireless let a few gears slip. That's unfortunate, because now they've set the expectation that the service isn't ready for prime time as a result. Reports of performance have been quite excellent on an unloaded network.
The problem? Computerworld reports that a handoff from 3G to LTE can take up to two minutes. A spokesperson told the reporter, "Hand-offs can take up to a couple minutes, but that was expected and a fix is in the works."
If it simply were an inherent problem, that's one thing. But it's clear this can be fixed in software, and is considered a bug. That makes it far less acceptable. In the olden days, products weren't shipped broadly until bugs that would frustrate your early adopting, high-paying customers were worked out. Bragging rights were more important here.
AT&T's CTO has a blog post indirectly critiquing Verizon Wireless's early LTE launch: I pretty much agree entirely with this John Donovan post. Verizon's commitment to CDMA left it without a reasonable path to future higher speeds in 3G because Qualcomm's EVDO path wasn't compelling enough, and Verizon clearly wanted the worldwide advantage of converging on GSM.
That leaves Verizon stuck at about 3 Mbps downstream with EVDO Rev. A. Verizon Wireless clearly and testably has the most robust and most thorough 2G and 3G network coverage in the US. That's still an advantage and will remain one on the voice side and for a large number of users for whom consistency is more important than speed.
But its early launch of LTE is driven by a need to have a higher speed number to push to businesses and consumers while AT&T and T-Mobile complete rolling out HSPA 7.2 and HSPA+ (21 Mbps), respectively. These evolutionary 3G HSPA flavors provide most of the advantage of first-generation LTE, including somewhat reduced latency, while preserving full backwards compatibility all the way down to GSM rates.
AT&T CTO is pushing the message that moving from LTE speeds to EVDO Rev. A rates will be jarring to customers in terms of what's possible. I agree. The difference is so huge that they are effectively different networks—this is a similar problem Clearwire and Sprint have with 3G/4G converged service plans.
However, Donovan doesn't mention the three other advantages of LTE: capacity, coverage, and latency. Higher bandwidth doesn't just mean that everyone gets greater speed; rather, it means that there's more potential to serve simultaneous users at greater speeds. That's often just as important as peak data rates. Coverage is a factor, because the 700 MHz networks can reach further and penetrate indoors better than 850, 1700, 1900, and 2100 MHz networks.
And latency is huge: lower latency makes networks appear faster because the time for each initial connection for every transaction is reduced. LTE promises very low latency, and HSPA delivers a decent part of that. Reduced latency equates to better video streaming, crisper phone calls, and more responsive Web browsing.
AT&T will benefit from the coverage and capacity issues, based on customer complaints, more than Verizon. But an early LTE deployment focused on speed doesn't provide the full picture of LTE's potential, and it hides the gap Verizon will have for at least three years, if not longer, between current 3G speeds and its LTE promise.
Update: Clearwire's chief commercial office weighs in with a swipe on Verizon's LTE pricing.
The 5–12 Mbps downstream 4G service will launch 5 December 2010 in 38 US markets and 60 airports: Verizon is still engaged in ridiculous pricing. The service will cost $50 per month for 5 GB or $80 per month for 10 GB of data transfer. Given that the cost per bit should be enormously cheaper for Verizon Wireless, and that they should be pricing this competitively with wired broadband carriers in the same market, that's absurd.
Clearwire's hybrid Sprint 3G/Clear 4G pricing makes much more sense. Unlimited usage on 4G Clear network, and same 5 GB limit on Sprint's home 3G EVDO network.
Carriers and ISPs continue to try to retain same limits even as services bump up faster. Comcast has the same 250 GB monthly usage cap on its cable service, whether you're at 15 Mbps or 100 Mbps.
LTE is required to serve next-generation mobile devices with streaming media, low latency, and heavy interactive use straining under CDMA 3G speeds today, although AT&T and T-Mobile move into faster HSPA rates alleviates that in part. But LTE will also become an alternative in some markets to fixed broadband, if Verizon offers sensible pricing.
You can check on which markets are covered at Verizon Wireless's 4G coverage map. I'm hoping to get review gear to test, as Seattle is a launch market.
Clearwire is digging in: The company, majority owned by Sprint, is shaving expenses. This doesn't bode well. With aggressive competition for 4G services from AT&T and Verizon Wireless, cutting back seems to make less sense than trying to double down. Clearwire is laying off 15 percent of its staff and delaying new markets and handsets.
Clearwire had already said it was testing LTE, the alternative to WiMax. WiMax's chief advantage was that it was available long before production LTE gear, and could take advantage of broad channels that Clearwire and Sprint had available in spectrum they'd acquired. LTE is now coming to market, and will be the dominant 4G flavor worldwide, while WiMax has developed into a useful niche technology that could retain double-digit marketshare even when LTE is the powerhouse.
However, how can Clearwire redeploy in the middle of a cash crunch? Especially with $2b in debt and other obligations becoming due in 2011, as Stacey Higginbotham reports.
Sprint executives leave Clearwire board: I missed this story of a week ago until sensationalist headlines started to appear reading more into the tea leaves than the murky water at the bottom of the cup. Sprint is the majority owner of Clearwire. That remains the same. Sprint has the power to appoint seven of the 13 board positions. That remains the same. Four of the seven board members appointed by Sprint remain in place, and Sprint will nominate three replacements for its executives.
The issue appears to be concern over whether Clearwire and Sprint's business objectives diverge sufficiently that Sprint executives or employees would make decisions that could be construed as anticompetitive or even detrimental to Clearwire shareholders. The SEC, Congress, and other forces have been pushing since the Bush administration to require boards made up of independent directors who have little stake in the current management of a firm.
The issue of whether this signals a partnership with T-Mobile or dropping WiMax seems secondary to the larger antitrust problem. And, if the company were considering partnering with firms other than Sprint for ventures, Sprint employees would be required to recuse themselves constantly, which is awkward for governance. Google CEO Eric Schmidt left the Apple board in part because of the growing competition between Apple and Google, and how often he had to leave the room while business was discussed.
Whether WiMax continues to be Clearwire's 4G flavor of choice won't be directly decided by this move. Sprint and Clearwire still need to prove billions of dollars invested to buy spectrum and build out a network weren't for naught. The widespread affordable availability of LTE is still two or more years away. Verizon might be launching dozens of LTE markets this year, but the gear will be 1.0, power hungry, immature, and in limited varieties. Clearwire can't afford to wait on LTE, and will still be deploying WiMax even if it makes a technology decision to switch to LTE in the future.
InfoWorld's Galen Gruman makes a number of conclusions in his WiMax Is Now Likely to Die article that I disagree with. Sprint execs leaving the Clearwire board won't affect ownership of Clearwire by Sprint, nor the inter-tie contracts in place between the two for network use and roaming. However, the likelihood of all US carriers and most worldwide carriers converging on LTE seems ever more likely.
Virgin Mobile has upped the ante on cellular data: Despite being owned by Sprint Nextel, Virgin Mobile is challenging all four major US carriers with an as-you-need-it, no-contract $40 unlimited 3G data plan. The plan lasts for 30 days. Virgin previously had four levels of service topping out at 5 GB for $60 used within 30 days. The new tiers are $10 for 100 MB over 10 days or $40 for unlimited data during a 30-day period.
Because Virgin Mobile also offers the MiFi cellular router for a low price ($150, no commitment), it now has a killer offering. Use a MiFi with an unlimited plan and avoid the overage fees or throttling from every other competitor.
This also guts tethering plans. I'm an AT&T customer with an iPhone 4, and I also own a 3G iPad (with no current active service plan). I typically now travel with the iPad and activate a plan on the road. I had figured on my next trip in which I needed a laptop, I would switch to tethering on my iPhone 4 (from a $15/200 MB plan to a required $25/2 GB plan plus $20 for tethering). That now seems unappealing.
Instead, I should pay the $150 for the Virgin Mobile MiFi, and pay $40 whenever I'm traveling. Then my iPhone and laptop can both use Wi-Fi to access Sprint's 3G network, and if I'm traveling with colleagues, I can share access with them as well.
Sprint recently dropped its MiFi offering (so far as I can tell) in favor of the Overdrive 3G/4G, which works on its Clearwire division's 4G WiMax network (no limits on use) and the 3G CDMA network with a 5 GB cap. (It's $350 upfront or $100 with a two-year contract at $60/mo.) You can also go to Clearwire and buy a similar product (the Spot 4G+) with a $55/mo service plan for the same terms.
GigaOm notes that Clearwire's 201 target of 120m people covered in 2010 could be hard to reach given its 51m passed numbers today: Stacey Higginbotham writes that Verizon's expected 2010 launch of a 100m-passed LTE 4G network could put a crimp in Clearwire's plans. However, as she notes, Clearwire's 4G pricing and limits (none on most plans) could provide an advantage over AT&T and Verizon, which place relatively tight limits on mobile broadband today.
Clearwire adds three products, including integrated mobile hotspots: The new Clear Spot 4G and 4G+ are mobile hotspots that don't require a separate USB modem, as with their predecessors. This increased cost, but also flexibility, as you could use the USB modem either with a laptop or the router.
The Spot 4G is $100 or $5/mo on lease, and connects up to eight devices over Wi-Fi. It works only in WiMax coverage areas. The 4G+ is $225 or $6/mo on lease, shares itself with up to five Wi-Fi devices, and works over Sprint's 3G EVDO and 4G WiMax networks. Both routers will ship in July.
Clearwire also improves options for Mac users with a new 3G/4G hybrid modem, the 4G+ Mobile USB Series S ($115 or $6/mo on lease), which works with Mac OS X 10.5 and 10.6, as well as Windows XP SP2, Vista, and Windows 7. It ships 1 July 2010.
4G usage is unlimited. 3G usage has the usual restrictions: 5 GB per month, overage fees above that, and no more than 300 MB per month on roaming (read: Verizon Wireless) networks.
Walt Mossberg reviews the Sprint Evo 4G: The new phone works on 3G and 4G WiMax networks, and Mossberg tested it in Sprint's oldest WiMax market, Baltimore, and in a new on in D.C. While he notes that he saw the fastest cell performance ever, that still wasn't saying much--he found speeds to be far below the maximum.
Battery life is the real issue. With 4G active, he couldn't get through a day on a full charge. Storage for third-party apps and data is strange. Mossberg writes that there's just 400 MB out of 9 GB available. The phone runs Android, and has a Wi-Fi hotspot feature and tethering available ($30 more per month).
The mouse roared, the elephant lumbers forth: T-Mobile, as the distant fourth-largest carrier in the US by subscribers and revenue, needs to have a gimmick. It chose to use its limited 3G network allocation to push for the fastest possible flavor available right now. The firm went to HSPA 7.2 network wide (including backhaul support) before AT&T; AT&T is using HSPA 7.2 but says it hasn't lit it up nationally due to backhaul issues.
T-Mobile several months ago put HSPA+ (21 Mbps raw downstream) on the table, and early this year said it would have the coasts deployed by summer, most of the rest of its coverage area (which is approaching AT&T's and Sprint's) by the end of 2010. AT&T seems stuck on its HSPA 7.2 upgrade with LTE looming as the next network refresh. AT&T even said last year it wouldn't deploy HSPA+.
I guess the future revenue of LTE seems too far in the future, and AT&'s CEO said on Friday (official press release not yet on its site) that it would roll out HSPA+ apparently the 14.4 Mbps flavor) on its US network to cover locations serving 250m people by year's end.
Sprint covers north of 230m people with 3G; Verizon somewhere over 280m; T-Mobile's 3G network passes more than 200m at last check, but it has licenses to allow coverage as large as Verizon (whether that's a good return on investment is a separate question).
Verizon continues to have the short stick in this 3G/3.5G/4G competition. It's EVDO Rev. A service tops out at 3.1 Mbps, and while it will have a super-fast LTE network running with several Mbps downstream service in late 2010, comparable coverage to T-Mobile or AT&T's HSPA+ network won't come until perhaps 2012. Sprint comes in third with Clearwire planning just 120m people covered with 4G WiMax service in 2010.
Clearwire has changed terms of a deal with Intel that required support of WiMax through late 2011: GigaOm reports that on the call, Clearwire's CFO said that the prior contract required Clearwire to use WiMax through 28 November 2011. Now, either party can back out with 30 days' notice on that commitment.
Intel has driven WiMax to its current state after apparently having dissatisfaction in the early 2000s with the direction, nature, and speed of cell carriers' drive towards mobile broadband. Without Intel, WiMax wouldn't have proceeded in the US, and perhaps would have stalled worldwide.
At the time of Intel's first real commitment, 3G networks were slow and hardly built out in much of the world, and the roadmap for faster speeds wasn't fully in place. A CDMA and GSM battle was still underway in the US, and LTE was still on the drawing boards.
The 2010 landscape is entirely different. AT&T and Verizon have the spectrum to deploy LTE and the commitment. LTE is being piloted all over, and is moving into production.
While hundreds of WiMax networks are in place worldwide--in part due to its availability, relative ease, and fairly wide range of equipment across many bands and encodings--most carriers plan a migration to LTE. In fact, WiMax in its next version and LTE may not be that far apart in terms of the underlying technology, just the encoding mechanisms.
Intel, Google, some cable giants, and others have a big investment in the Clearwire portion of Sprint (as does Sprint itself), while Motorola, Samsung, Zyxel, and others have poured billions into hardware development for consumer and carrier gear. But if it's not going to wind up being cost effective or competitive, a transition to LTE makes more sense than spending billions more heedlessly.
With a transformed market, Intel may be less concerned about bandwidth and content control, walled gardens, limited speeds, and other factors that led it down this decade-long path.
The biggest problem I see with Clearwire switching to LTE is that while Clearwire has extensive 2.5 GHz holdings that would let them deploy nationally with high data rates, the 2.5 GHz range requires about four times as many base stations as AT&T and Verizon's 700 MHz range to provide the same footprint, while penetrating indoors more poorly.
It may be that less equipment is needed at each location in 2.5 GHz because more gear is spread out than for 700 MHz towers, but securing real estate and operating four times as many nodes introduces fixed costs above any savings in having less gear in each place.
Fascinating development. Were Clearwire to exit WiMax, that would put a pall over the issue of producing devices in enough quantity for that technology.
It won't ship til summer, but Sprint has a 4G phone for the Clearwire network: The HTC Evo 4G will be the first WiMax smartphone--first mobile WiMax phone of any kind that I'm aware of--and Sprint plans to promote it as an early entrant and fast alternative to competing networks. The Android-based phone will receive data at up to 6 Mbps, the rate Clearwire promises in its WiMax-covered cities. (Clearwire says typical downstream rates are 3 to 6 Mbps, with 10 Mbps bursts.)
The Android-based phone will work on 3G CDMA and 4G WiMax networks, and is designed for video. YouTube, for instance, will kick into a higher-quality mode showing video on the 4.3-in screen when the phone is on a 4G network. The phone has a kick-stand for hands-free viewing, and HD video output for a big screen. The phone also will have full Flash support; it must be relying on the Flash 10.1 release due out later this year.
The Evo will also be a hotspot-in-a-phone, sharing a 4G connection with up to eight other devices over Wi-Fi.
A key advantage of Clearwire's WiMax network is a lack of bandwidth limits: As carriers boost speeds on mobile broadband networks, users will increasingly employ those networks like wired broadband--and run into the 5 GB limit beyond which insane per MB fees are charged--from 5 cents to 20 cents a megabyte, which is $50 to $200 per gigabyte. Clearwire sees its freedom from limits as an advantage, quite clearly.
Clearwire's CEO Bill Morrow said in a press release that the average 4G customer on the Clear network uses over 7 GB per month. That would cost a 3G user $100 to $400 extra, but it's included in a $45/mo unlimited 4G mobile plan (or cheaper for fixed home service); unlimited data is also included in all home plans (from $25 to $45/mo). A home and mobile bundle is $65/mo.
AT&T, T-Mobile, and Verizon will be under increasing pressure in markets that Clearwire enters and co-markets with Sprint (which can offer 3G back-up in the rest of the country) and cable operators like Comcast. I would suspect that if Clearwire's service tests out well, heavy users likely to be long-term subscribers and pay regular overages will simply move to a Sprint/Clearwire plan, using a 3G/4G combined modem or portable router.
On the issue of capacity, Clearwire said it is "doubling the number of transmitters and receivers per site, thereby boosting potential end user speeds by approximately 20-30 percent." This would answer a concern raised to me by Novarum's Ken Biba, who found in testing earlier this year that Clearwire wasn't dense enough in its deployments for its suggested download speeds. Clearwire also said it's increasing backhaul by 250 percent or more.
Update: Biba asked me (rightly) to clarify that adding gear and bandwidth at existing locations won't solve the speed issues he spotted, which require denser deployments (more nodes per square mile, for instance). Clearwire's bandwidth is all in 2.5 GHz, which needs relatively dense outdoor deployments even at the power levels Clearwire is allowed in that band.
Clearwire says it's on track for 120m people passed in 2010; beyond that, more money needed: paidContent.org runs down Clearwire's earnings news, with the firm bringing in $80m in revenue, but a loss of nearly $100m. Subscribers stand at 688,000. The company will spend about $3b in 2010 to get to its target of 120m people to whom service would be available. But if the network is to grow any bigger, more cash is needed.
In contrast, T-Mobile has crossed 200m already in upgrading its 2G network to add 3G, and it's ostensibly on track to something approaching AT&T (233m) or Verizon (280m) levels. (Yes, those Christmas ads in which Verizon said it has the biggest 3G coverage are correctly, but the 50m gap is in small towns; AT&T has the big markets.)
Clearwire is leaning on Sprint for its fallback position, Sprint being the majority owner of Clearwire. If you get a combo 3G/4G USB modem or portable 3G/4G-to-Wi-Fi gateway, you can use 3G (up to 5 GB on Sprint's network) when away from a Clearwire area, and 4G in the footprint.
But Sprint has underspent on 3G. The company reports (the latest number I could find) 270m people passed with 3G with roaming. That's a critical point. Compare Verizon's native coverage map with Sprint's map, where Sprint shows gray for roaming, orange for 3G and tan for 2G:
And here's Verizon's, where dark blue is 3G, green is 2G, and yellowish is roaming:
Sprint has the big cities, sure, but it doesn't disclose its non-roaming footprint, and it limits 3G customers to no more than 300 MB per month of roaming data, after which point it can choose to cancel a mobile broadband contract.
Sprint will sell its Overdrive, a router that handles 3G and WiMax, starting 10 January: The device, $100 with a two-year contract, works with either 3G CDMA or 4G WiMax, preferring the 4G network. It's $60 per month, which includes unlimited WiMax usage and 5 GB of 3G. The device appears to be nomadic, rather than mobile: that is, intended to be taken from place to place, but not used while in motion. (It's unclear whether or not there's a car adapter included.)
From a carrier with no 3G offerings 18 months ago, T-Mobile has turned the ship fast--and turned the table on its competitors: T-Mobile used today's announcement of a new 3G USB modem to lay out its aggressive plans for 7.2 Mbps HSPA and 21 Mbps HSPA+ deployment nationwide.
Starting from no customers in second quarter 2008 and clutching a handful of 3G spectrum, the firm now covers 240 cities and passes 170m people. T-Mobile's Jeremy Korst, director of broadband products and services, said in an interview that the number will hit 200m by the end of 2009, which covers nearly all the major urban areas. By contrast, Clearwire plans coverage of 120m people with its Wimax service by the end of 2010.
But perhaps more important is that T-Mobile will have 7.2 HSPA, which runs at a raw downstream data rate of 7.2 Mbps, on all its 3G nodes by year's end. On the upstream side, T-Mobile will gradually upgrade to 2 Mbps starting in early 2010.
This contrasts with AT&T's previously announced but much more moderately paced plan that gradually upgrades the current, seemingly overloaded 3.6 HSPA network to 7.2 HSPA through the end of 2011, at which point AT&T will still have only 90-percent 7.2 HSPA on its 3G network. By the end of 2010, only 25 of 30 major markets will have the faster HSPA flavor, the company has said.
The bigger news, though, is that T-Mobile is going full-court press on HSPA+, a 21 Mbps flavor already deployed by several carriers worldwide, and which T-Mobile launched for test purposes in Philadelphia in September. The company will start rolling out HSPA+ in 2010 on a "fairly broad-scale" basis, Korst said.