How will you be computing in the next 12 months? PC Authority brings you 2005's big technology stories - before they happen.
This is it. We've come to the time of the year when we make fearless predictions about the year ahead, and quietly cross our fingers and hope such predictions won't come back to bite us in the butt a year from now. Over the next few pages we'll walk you through the products that we think will get most of the attention this year, but before you read, we have two words for you: caveat emptor. The IT industry is a fickle one, and nearly every year the big story turns out to be, much like the Spanish Inquisition, something that nobody expects. Last year it was the iPod and iTunes. What surprises will 2005 bring?
A year in review
Let's face it, 2004 was a bit of a yawn on the technology front. Advancements in PC power stalled, with the possible exception of graphics cards, and new technology innovations were rare in both the business and consumer space.
|
Other cool stuff from 2004
We've mentioned the iPod and broadband, but there were a few other notable happenings in 2004. They may not have set the market on fire, but here's our pick of some of the best things to happen in the last year.
Cheap colour laser printers - colour laser printers have dropped so much in price that their purchase cost is now challenging inkjets, and their cost per page blows inkjets away. Two years ago, it would have been unthinkable to get a colour laser for $600. Now, just about every printer vendor has a product at that price.
Mozilla - it still only holds a small proportion of the browser market, but Mozilla started making an impact on the total market dominance of IE. For good reason too - it's a better browser, and doesn't have the gaping security problems of the Microsoft browser.
Web services - finally everybody has agreed on the method by which applications should talk to each other over the internet. We have high hopes for web services, and expect it to drive a new wave of innovative integrated applications in 2005 and beyond. |
If we were to pick the two big technology stories of the last year, it would have to be the final arrival of broadband and the enormous success of the Apple iPod and the iTunes service.
|
|
| Apple's iPod had a massive year, and contiual updates and third party add-ons suggest it's set for another big one. |
Broadband finally took off because of price cuts in the first quarter. Telstra BigPond announced sub-$30 price plans, and its parent company was forced by the ACCC soon after to lower its wholesale prices so that other ISPs could compete with this offering. Suddenly, many Australians were looking at broadband anew, dumping their dial-up accounts and picking up low-cost services for their home and business. In August, the number of broadband connections in Australia reached a million, and is expected to grow to three million and beyond in the next five years.
The growth in broadband spurred growth in home and small business networking. With a compelling reason to link their PCs up now, people were looking to Ethernet and wireless connectivity to share their new broadband connections. Wireless, much as in 2003, was a big winner in 2004.
Telstra's iron lock on the 'last mile' of internet connectivity in Australia was also challenged in 2004, with the introduction of wireless broadband and third-party DSL provision. So much so, that we're predicting that wireless broadband is going to be one of the very big stories of 2005. If there was any technology that struck a chord in 2004, however, it was the Apple iPod. You know you've struck gold when the general public refers to any similar device by your product's name -- it's a position that the iPod shares with other products like Kleenex and BandAid.
 |
| Apple's iTunes music store will hit our shores in the next few months. |
Apple also made waves with iTunes, an online sales site for music. After years of record companies trying to sue the pants off anybody even tangentially involved in distributing music online, Apple showed that, given a reasonable price, people are prepared to pay for music they acquire online. It was refreshing to see, and is hopefully the beginning of a new era of media distribution. Now we have to wait to see if the movie studios get on board.
VoIP
At the end of 2004, using the internet infrastructure to make voice calls was about the only thing that anybody wanted to talk about. Voice over IP (VoIP) hype had reached critical mass, and the offers of cheaper phone calls came thick and fast. Ultimately, that's what VoIP is all about. In most cases, it doesn't offer any significant technical or service improvements over the existing public switched telephone network, and in fact has some serious problems, such as the relatively poor reliability of the internet, variable service levels and the inability to make emergency phone calls when the power goes down.
But still, the chance to call anywhere in the world for free, or at least at local call rates, is enormously compelling. Will it fly in 2005? It's a distinct possibility. But it was also possible in 1997, when OzEmail launched OzEmail Phone, a VoIP service that ultimately failed. The technology has improved since then, however, and the relentless hype has at least made people aware of the possibility.
 |
| Hardware that bridges the gap between your existing phone and your internet connection are a good way to get into VoIP. |
Many PC Authority readers will probably think of voice over IP in terms of Microsoft Messenger/NetMeeting or Skype - a way to bypass tolls by using a 'softphone' application on your PC to talk to somebody else using the same application. Softphones are great, allowing effectively free voice calls to anywhere in the world, but the person at the other end has to be sitting at their PC and using a compatible softphone themselves for it to work. With Microsoft Messenger, you already have a working softphone built into Windows (just click on 'Start Talking' on the right of the chat window), and all you need is a $25 headset -- the same kind used for voice communications in multiplayer gaming -- to make it happen.
Service providers are interested in VoIP as a mechanism for entering the very lucrative telecommunications business without having to face the astronomical costs of building a telephone infrastructure. The well-founded belief is that the mass market doesn't really want to change its telephony practices, but is always up for cheaper stuff. Service providers plan to offer services that work, from a user perspective, exactly like regular phones. You pick up the handset, dial a number, and talk. It's only in the back end that VoIP does its thing, routing the voice messages through the internet and back through a voice gateway at the other end, bypassing the regular telephone infrastructure.
The key challenge for this kind of service is the interworking of VoIP and PSTN services, including phone numbers, so that people can call non-VoIP users and vice versa. Fortunately, the wide acceptance of SIP (session initiation protocol) and standards-based compression means that calls between different VoIP services and devices will generally work well.
 |
| We'll start seeing fully PC-based phones hitting the consumer market this year, as well as basic handsets that interface directly with the PC. |
Such services exist today. They usually require an internet connected box that your telephone handset plugs into, or a special IP phone (which looks and works just like a regular phone). These boxes and phones communicate over your broadband connection to a service provider, which routs the voice call over the internet to a local telephony gateway, which then puts the call on the PSTN. So if, for instance, you were calling Los Angeles, your call would be routed over the internet to an exchange in LA, where it would be put on the PSTN there (at the cost of a local call) and routed to its destination.
The problem with these services right now is the limited number of internet telephony gateways. The services work well and are very cheap compared to regular international phone calls, but aren't that effective if you want to call somewhere that lacks a gateway.
The other big area of interest for VoIP is for business PBXs, often linked with an external VoIP phone service. IP PBXs allow businesses to unify their phone and data networks, and to link their office PBXs together over virtual private networks. Calling the Melbourne office from the Sydney office over the IP PBX, for instance, would be effectively free (excluding data volume charges for internet access).
It's in this area VoIP is most likely to make its biggest splash in 2005, which will also hit the small and home offices. We expect that products like the Cisco Integrated Services Router will do big business this year. These devices provide both regular IP routing as well as act as an IP PBX and PSTN gateway. Just plug a set of IP phones into the Ethernet line connected to the router, and you have yourself a very feature-rich PBX - including voice messages, call waiting, conferencing and other features. Through the PSTN gateway, calls can be made to regular telephones outside the company. It's very compelling, not that costly to implement (although IP handsets are still too expensive, in the $100 per phone range) and likely to be the real driving force behind VoIP in 2005. But, if you're reading this magazine, you're one of those that'll be interested in this right now. Companies like Engin and AstraTEL offer hard phones that plug straight into your broadband modem and so long as you have a connection up and running, you can make calls to regular landline phones at considerably reduced rates.
Wireless broadband
Midway through 2004, we saw the launch of some of the most innovative broadband services seen since ADSL was introduced. Led by Unwired and iBurst, wireless broadband became a reality last year (click here for a head-to-head review). This isn't your garden variety 'hot spot' Wi-Fi broadband, limited to airports and McDonalds restaurants, but wireless broadband from your home, at speeds and prices comparable to wired broadband.
The existing services are pretty compelling. Wherever your modem is, that's where your broadband connection is. You can take the modem to work with you, to a friend's place and back home at the end of the day. The service doesn't change.
 |
|
iBurst is available through many vendors and features notebook PCMCIA adaptors as well as the wireless access points pictured here. |
They're not going to replace the massively-overpriced general packet radio service (GPRS) for mobile phone net access soon, but they are a step forward. They might be challenged by 3G internet access services in the medium term, but we're putting our money on wireless broadband.
The coverage areas of the wireless broadband services are considerable, but largely limited to the major metropolitan areas of the state capitals. Unwired is, at this stage, Sydney only, although iBurst has a much greater coverage (iBurst does not, incidentally, sell services directly, but through resellers like OzEmail, Veritel, United IP and others).
Although the latency is a little higher than land lines, making wireless broadband less useful for gamers, the speeds offered by wireless broadband are very high - up to 1Mb/s downstream. Download volumes are a little more restricted than wired services, but not excessively so. In short, there are a few minor sacrifices to get wireless broadband, and one or two major advantages, especially if you like moving around.
In the long term, we expect these early, largely proprietary services to be replaced by WiMAX (or IEEE 802.16) services - that is, if the reality lives up to the specs. WiMAX is a standards-based protocol that can theoretically deliver up to 75MB/s within a 50km range of the base tower. Practically, it's expected that a WiMAX station will have a range of 5 - 10km.
WiMAX, if it happens, will mean non-proprietary (and thus cheaper) access equipment, faster speeds and more competition. It will have a low price of entry for service providers, with a few towers being able to cover most of the population of major metropolitan centres. WiMAX is not likely to happen in 2005, and may never happen at all if the technology doesn't work well in the field. Some serious testing and product design is needed before WiMAX hits the airwaves, so for now we're stuck on 1Mb/s wireless broadband.
|
The marginalisation of Telstra
It's not just wireless that's literally cutting Telstra out of the loop. Wired broadband providers are making big strides to eliminate the 'Telstra factor' from their broadband offerings.
As of now, nearly all broadband connections in Australia use a Telstra service at some point along the line. You may use a third party ADSL service, but it's very likely that the provider has to pay Telstra a very significant chunk of their service fee for the rental of the 'last mile' of the service to your home - the link between the exchange and your house. They also have to stick to Telstra's pre-set speed plans - 256/64Kb/s, 512/128Kb/s or 1.5Mb/s/256Kb/s - because these are the only options available. Telstra has no plans to make faster services available anytime soon.
The good news is that a number of ADSL service providers have taken matters into their own hands. iiNet, Internode, iPrimus and a few others are installing their own DSLAMs (DSL access multiplexers) into telephone exchanges around the country, with quite a few exchanges already installed. These DSLAMs allow the service providers to avoid the 'Telstra tax' and provide speeds not supported by Telstra. Already iiNet, for instance, allows users connected to exchanges into which it has installed DSLAMs to connect with upload speeds as fast as their line conditions can handle. On an iiNet DSLAM, a 512Kb/s download service generally also offers 512Kb/s upstream.
As a result of the installation of non-Telstra DSLAMs in the next year, we're going to see affordable broadband services of 2Mb/s and possibly even up to 8Mb/s offered. That is, if you're lucky enough to be on an exchange which has been installed. |
Multi-core processors
The processor business has been uninspiring of late. Processor speeds are maxing out, with the existing architectures starting to hit their physical limits in terms of clock speeds. But, late next year, processors are set to receive a shot in the arm, when a spate of multi-core processors hit the store shelves. Just about every major processor manufacturer has one or more designs in the pipe, and quite a few multi-core processors have already been shown.
A multi-core processor is effectively a single chip that contains several processing cores. Unlike Intel's Hyper-Threading, which uses software tricks to make it look like a single core is more than one processor in order to take best advantage of the various functional units in the processor, a multi-core processor has several physical processing cores.
So, cutting out the tech jargon, a multi-core processor is two (or more) processors in one. They will appear, to the operating system, as two processors. They act as two processors and for all intents are purposes are two processors, but located on a single piece of silicon. Apart from the cost advantages, multi-core processors allow cores to share common components (such as cache) and use high-speed communications not possible in multi-processor systems.
 |
| The die shot of the Opteron (top) shows a single core on the left addressing the cache on the right. Compare this to the shot above, which shows the two separate cores connecting to a large cache. |
From a practical standpoint, a multi-core processor, much like multiple processors, will not make your computer go twice as fast. It's more accurate to say that the processor can do twice as much at the same speed. You'll really notice the value of multi-core processors when you're doing something that normally sucks up the entire resources of your computer, such as rendering a video. With a single core processor, the computer becomes sluggish and unresponsive, but when you have multiple cores, you still have one or more free cores to do the other things you want to do.
Once multi-core processors have been around for a while, it's likely that we'll see applications designed to really take advantage of the extra cores. The operating system applies each processing thread to a single core, and today, most applications create a single thread, although more complicated and better programmed applications can sometime use multiple threads. Thus, a one-thread program can only run on a single processor. In the future we will see more multithreaded programs -- a game, for instance, may use one thread for graphics, one for audio and one for everything else. The operating system can divide these threads out to the different cores.
The key question, of course, is when and how much. We can expect that multi-core processors are not going to appear for another six months, and will most likely appear first for server platforms like AMD's Opteron and Intel's Xeon and Itanium.
AMD is likely to be the first of the X86 vendors to hit the market with a dual-core processor. In August 2004, AMD showed off a working AMD Opteron dual-core processor, and said that it plans to release the product midway through this year. The processor was based on the existing AMD64 architecture (and includes a limited set of SSE3 extensions), and is compatible with existing 940-pin Opteron motherboards and memory. It hasn't yet announced any consumer versions of the dual-core processor.
At the time of writing, Intel had yet to announce anything specific, but it does have multi-core processors on its roadmap for the end of 2005. It showed off its first dual-core processor in September, a week after AMD's demonstration. It will first hit with its Montecito dual-core Itanium, and has plans to extend the line to a 16-core Itanium. At the consumer end, the story on dual cores is unclear. We expect, however, that with AMD rapidly bringing dual-cores to market, Intel won't be far behind. Intel also has the advantage of having a 90nm manufacturing process that's up and running successfully (it's using it to manufacture its Prescott processors), which gives it a distinct advantage from a manufacturing standpoint. The design work involved in building dual-core processors should not be that great, and we would not be surprised if it abandons the Pentium 4 architecture for its dual core processors and instead uses the powerful Pentium M, which is based on a modified version of the Pentium 3 architecture.
IBM has stolen a march on both companies when it comes to multi-core processors (as have high-end vendors like Sun and HP). Its existing Power5 processor already has multiple cores, and its upcoming Cell processor, developed with Sony and Toshiba, will use an advanced multi-core-like architecture.
Two processors for the price of one - sounds like a sweet deal, doesn't it? Of course, we haven't seen pricing for the new chips, and we expect the price tags for the first generation to cause severe sticker shock. Still, multi-core processors will be worth the wait.
Blue laser storage
Blue laser storage. It's not very catchy, but it sums up one of the biggest technology battles of 2005. Blue laser storage, in the guise of Blu-ray and HD-DVD, represents the next step up from DVD for optical storage. In both cases, the specs are set, the demo products are out there and the various factions have settled on their choice of technology. Much like the DVD+R/DVD-R schism, Blu-ray and HD-DVD are splitting the manufacturers of optical drives and media into camps.
Both technologies use short wavelength blue lasers, which allow very high resolution pickup of information on the disc. In the case of Blu-Ray, this translates into a capacity of up to 27GB on a single layer disc, and 54GB for dual layer discs. HD-DVD discs have capacity for 15GB of data per layer and support for two layers, and are allegedly cheaper to manufacture than Blu-ray discs.
 |
| The big competitor to Blu-ray this year willl be HD DVD. It's also based on blue laser, which allows it to access tightly packed discs, and therefore greater capacities. |
Both formats were designed for the storage of high-definition movies, and a full set of specifications exists for consumer players. There are also write-once and re-writable versions of each specification, just like DVD.
Recorders and players for both formats are already available, though far from affordable, and no movies have yet been released for either format, in either high definition or standard definition.
As has happened with just about every optical media standard since the CD-ROM, the standards battle will inevitably slow down the pace of acceptance of blue laser technology.
We can't pick a winner yet, nor can anybody else. In one corner you have Sony leading the Blu-ray camp, with plans to use Blu-ray in the Playstation3, likely to be released at the end of 2005 or in early 2006. In the other, you have the DVD Forum, the main DVD body, pushing HD-DVD. The movie studios are divided about which format they will release their high definition movies on. Well, to be more accurate, the studios that Sony owns (Sony Pictures and Metro-Goldwyn-Mayer) have come out in support of Blu-ray, while Warner, Paramount, Universal and New Line have announced support for HD-DVD. HD-DVD also appears likely to be the disc format for the Xbox, which is expected to come out at around the same time as the Playstation3. If we were to make a pick now, we'd say that HD-DVD has the upper hand, but we wouldn't recommend forking over your money just yet.Media servers
Our earlier potshots at Windows Media Center notwithstanding, 2005 could well be the year of the generic media centre. Already we've seen significant penetration into the homes of hobbyists and the wealthy (the usual early adopters), and we expect to see a lot of good products hit the shelves this year.
A media or content server is fundamentally a device that divvies out movies, music and images to terminals around the home or office. Those terminals could be TV screens, LCDs, PCs, gaming consoles, set-top boxes, speakers or pretty much any device capable of outputting media. From a user-perspective, they work pretty much like the pay-per-view systems you find in hotels, but without the paying. You own the media, you rip it to the media server, which is networked to the various devices around your home or office. With a remote, you select the movies you want, and it plays them wherever you are.
It's perfectly possible right now to build a media server (it is, after all, just a file server at heart) and use small PCs or even modified Xboxes running Xbox Media Center (a homebrew open source application) connected to TVs around the house as the players for the media stored on the server.
To bring media servers to the masses, however, various vendors, including PC suppliers using Windows Media Centre Edition, are building custom boxes, designed to store massive amounts of media with easy access from devices around the home. At the very expensive end, you have products like the AMX MAX, which can have up to 7.2 terabytes of media stored and up to 25 movies being played back at different locations simultaneously. It can also draw movie information from the internet and even bookmark movies, so you can stop watching a movie in one room and pick up where you left off in another. The server distributes media to small black boxes which have TV and VGA-out connectors for attachment to screens. Sadly, prices for the AMX MAX start at US$10,000.
For the rest of us, there are Windows Media Servers, which will likely be able to use Xboxes or the next version of Xbox, as media terminals. There are also dedicated servers from Cisco, Denon and others that will cost a few thousand dollars.
The other thing we expect to see in 2005 is serious attempts to introduce personal video recorders to the Australian market. PVRs use hard discs to store recorded television shows, in some cases hundreds of hours of it. With a PVR, you can 'pause' and 'rewind' live TV, as the PVR records whatever television show you are watching, and buffers it if you decide to pause the live TV.
The key to the success of PVRs in Australia will be the creation of a proper electronic program guide. Previous releases of PVR products in Australia from Samsung and now Microsoft OEMs have been hamstrung by a lack of an EPG. In the US, the success of the TiVo and ReplayTV PVR products has been linked to the simplicity of the online guides. If you wanted to record every episode of American Idol, you could just select it from a menu and the PVR would do the rest. When you're ready, you go over to your TiVo and all the episodes are there, ready for viewing.
All that said, we wouldn't call media centres and PVRs a lock in the Australian market. They represent interesting opportunities, and we expect there will be a good deal of hype, but the success of the products depend on the fickle consumer. But then, as we said earlier in the piece, there are no guarantees in the IT industry, and it's nearly always the products that we don't see coming that turn out to be the big success stories.
The Maybes
There are some technologies that we think have potential, but we're not quite sure will take off just yet. Here are a few:
Legitimate downloadable movies - they did it with iTunes, why not with movies as well? We're hoping that the movie studios will stop trying to sue everybody and start allowing people to download movies legally at a reasonable price.
The success of the pay-per-download music site iTunes makes it more likely that movie studios will start thinking about providing legal downloads, but we haven't yet seen much movement in that direction. If it does happen, expect Apple and Microsoft to be at the forefront.
RFID - radio frequency identification (RFID) almost made it onto our list ahead of media servers. RFID comes in the form of tiny tags that attach to objects that wirelessly provide electronic information about the object. RFID is already being put to use for inventory tracking in warehouses and in a few stores. It tells the owners where the products are, how many are on a shelf and can be used as an alternative to barcode readers. The range of RFID varies, but is generally 2 - 10 metres.
In the future, its potential is amazing. Imagine packaged food that comes with an RFID tag that automatically tells your oven how to cook it. Or clothes with RFID tags that talk to your washing machine, informing it of the best washing mode to use.
But that's not for 2005. Right now, RFID tags are way too expensive to be practical for anything but the most expensive of items. They currently cost a few dollars per tag. When they get down to a few cents, then we'll be talking.
Home automation - much like media servers, home automation is largely a pastime of the hobbyist and the rich. Home automation is effectively a set of scripts or rules that control the behaviour of devices in the home. Turn the TV on, and the lights automatically go off and the curtains close without prompting. Air con and lighting automatically switch on when a motion sensor detects somebody in the room. Have the hot water heater turn of at a set time to save power. These are the core kinds of functions that home automation delivers.
CEBus and X10 have been around for some time, allowing devices in the home to communicate with and control each other, but they have never had much of an impact on the general consumer. Reports from resellers and installers of home automation equipment suggest, however, that home automation installations are skyrocketing. Some are reporting over 200 percent increases year on year in their residential businesses.
Fizzers
Some technologies are beyond all hope. In spite of our best wishes, here are a few that have failed to fly, and will probably continue to do so.
High definition television - We're going to go out on a limb here and fearlessly predict that in 2005, the price of high definition television sets is still going to be way too high, and high definition broadcasts way too rare, to register even a blip on anybody's interest charts. Hell, hardly anybody even cares enough about free to air standard definition digital to get enough momentum going to drive down prices on set-top boxes and sets. Although we like high-def, we don't like it enough to want to fork over $10,000 or more to buy a TV set that just barely meets the minimum qualifications for high-definition. Perhaps in 2006, acceptance of Blu-ray or HD-DVD will drive sales and innovation in HDTV, but we're not holding our breath.
3G - If we had run this feature in 2001, 3G data access would have almost surely been on our list. Hell, mobile phone vendors around the world had just forked out ludicrous amounts of money for 3G spectrum, and the world was buzzing with 3G hype. Australian companies paid $1.17 billion for spectrum in March 2001, with Telstra anteing up over $300 million.
Well, we would have been embarrassed. 3G data access hasn't happened, and we've all but given up hope of seeing an affordable service that lives up to expectations. Wireless broadband service providers such as iBurst have popped 3G's bubble, and WiMAX is likely to be the nail in the coffin of 3G data.
Windows Media Center Edition - Ever since IPC launched the My GNiE in the mid 90s, the loungeroom PC has been a holy quest for the PC industry. Gateway tried and failed. Toshiba tried and failed. Nobody could get the sales volumes high enough to justify the development expense. The loungeroom PC was limited to a very specific niche - and members of that niche are generally savvy enough to build their own systems.
People are thinking that maybe this time it will be different. It's Microsoft, and they've got enough money to make anything happen. In our November issue we pulled apart Media Center, and saw the future of loungeroom entertainment. Unfortunately, in our December issue we looked at how it was going to be implemented in Australia and we had a different story to tell -- no electronic programming guide. This is Microsoft's killer app, and Microsoft's MCE PCs will stay on the shelves until the EPG is implemented.