Qualcomm announced the ‘first’ smartbook [*1] yesterday. It’s due in 2010 and it’s going to be a very very tough challenge to make an impact of this wildly dynamic mobile device market we’re seeing right now. My opinion is that ‘smartbooks’ are not going to impress anyone in the first half of 2010 but as we near 2011, they have a chance to break through. Operating system and processing power are the key issues but in this report I talk through a number of the ingredients that go together to make the complete smartbook product.
OS. There isn’t one, yet. A hacked Ubuntu ARM build, Android or Windows CE with overlay software just isn’t going to cut it in the mainstream market. Interestingly enough, Nokia’s Maemo is probably one of the best bets for a smartbook OS right now although Google could do something with Chrome OS. I have my doubts about Chrome though because consumer operating systems need to be very dynamic. How are Google going to handle the distribution and monetisation of browser apps? Android is another possibility but the main branch of that operating system will prbobably stay focused on smartphones meaning that a ot of work has to be done to build a new version and to keep it syncronised to the main version. The OS problem will be solved but I’m not expecting anything too thrilling in 2010.
Speed. Smartbooks will be slow. When we look at the top smartphones on the best ARM-based platforms we see a fast smartphone. Take that experience and put it in a laptop format and it doesn’t impress. I really love the speed of my 800Mhz TI OMAP 3430-based Archos Android one of the most powerful ARM-based devices on the market but it would be a slow slow experience if I was to try and do laptop-like work on it. It’s like cheap 12 inch netbooks masquerading as notebooks. They look great, have a good price but don’t give the end user an experience they can call thrilling.
The fact is that even the latest ARM cores can’t compete with Intel’s netbook platform at general computing activities. Crunching a webpage needs a strong, general purpose processor capable of multi-tasking. Look to the multi-core Cortex A9 CPUs that will appear in late 2010 for a web experience that will be similar to 2008/2009 netbooks. Of course at that point, Intel will also be offering their next platform.
Battery life. You’ll hear a lot of talk about battery life and unfortunately it will be very difficult for people to understand what the figures mean. In the netbook world you can usually knock 30% off the marketing battery life figures and get a reasonable idea of how the real-world battery will be. In the ARM-based world where the difference between ‘idle’ and ‘off’ is slim, the differential between long battery life and the all-important real-world battery life figure is huge. With the wifi on and screen off, a small smartbook could, like a mobile phone, sit happily for a few days. Turn on the 8 inch screen tand start browsing though and boom! You’ve just increase your battery drain by a multiple of 100 or more. Start browsing a few tabs and that smartbook won’t get a chance to idle. My only advice is to ignore ‘idle’ battery figures. [Read about ‘getting things done’ battery life and why a lower-power processor can increase system power usage here.] Battery life will be better as the ARM platforms are good at turning things off when not used but don’t expect miracles. Remember that while a smartbook might give you 50% more battery life, it might take you 50% longer to do things on it!
Size. Size is not really an argument for smartbooks. Coming from the ultra mobile PC world I know about size compromises. Smaller can be more mobile but when there’s typing to be done, nothing substitutes a near full-size keyboard which means a relatively wide layout. A 1024×600 screen needs space too. Going any smaller than an 8.9 inch screen means major compromises for people that want to use smartbooks for typing and browsing the web. Netbooks and Smartphones face the same industrial design challenges.
3G. The always-connected scenario is one element of smartbooks that many have been talking about. There are two problems here. 1) Power drain 2) Cost. It’s possible that the 3G is there to tempt carriers into channeling the device to customers which, based on the excitement I’ve seen over subsidised netbooks, might be a problem. In many parts of Europe users are switching to pay-as-you-go data plans and in the U.S, the data plans are simply too expensive. A purchase price that includes data might be interesting but there again, it will push the cost of the smartbooks up. The cost of the 3G module will also have an impact on the device. A module may only cost $10 or $20 but when the total bill of materials needs to be under $100, that doesn’t help!
Weight. Based on my size argument above, there’s not much advantage that a smartbook has over a netbook. Take a look at what was done with the Sony P-Series. It’s a full PC in 600gm. Intel’s Moorestown will make that even easier. I would expect a small reduction of 100gm in standard size netbooks based on a reduced battery and motherboard size but smartooks use the same radio, screen, storage and construction techniques that every netbook manufacturer has access too so the difference between a credit-card sized motherboard and a postcard-sized one is minimal.
Instant on. The smartphone platforms used by the smartbook manufacturers do have an advantage here. They can go to sleep (or even ‘active’ sleep) well. There is good potential for an always-connected device here and this is one area where I’ll put my hand up and say ‘me want.’ Smartphones using the same platforms just don’t have big enough batteries to offer this feature so I’d be interested in something that connects to Skype, Gtalk, Twitter and stays connected all day.
Video. Given the right operating system it’s possible to get some really nice video playback performance out of the smartphone platforms but don’t forget that there are netbook platforms that have similar, and in some cases, exactly the same (ref: Intel GMA500) 3D and video-decoding capability. The issue is software, drivers, OS architecture and licensed codecs and that’s not an area where the smartbook manufacturers have exclusive control. Again, the same challenges occur here as for netbook manufacturers.
If smartbooks are done well they will be lightweight, attractive, value-for-money devices that will be a joy to market but just because they run on the ARM platform doesn’t mean they have an automatic all-round advantage. Waiting until the next generation of ARM platforms and operating systems (and trade-show seasons) are ready is probably not the best option so I understand why the manufacturers need to start now. I just hope that those marketing teams keep things in perspective. Everyone is in the same boat right now and will have to continue to work on their radios, battery technology, screen technology, OS, applications and the developer community.
Related Article: What Moorestown means for consumers.
[*1] In Germany, Smartbook is a trademark of Smartbook.de
Chippy, great article!
But I disagree with the weight advantage
for example:
5″ Archos 5 IT – 200 gm, Viliv S5 – 450gm
7″ SmartQ7 – 450gm / Atom UMPC – 700gm
9″ Always Innovating Touchbook – 650 gm, Archos 9 – 800gm
Innovative Converged Devices announced Vega – 15″ 1,3kg slate!
Sony P is a very expensive product – there is now such ARM product to compare with
Moorestone will be in a year
I think the 200gm advatage is considerable if speaking about sub 1,3 kg devices
I filmed a $80 Smartbook at http://techvideoblog.com/reviews/80-android-laptop-menq-easypc-e790/
It’s only ARM9 processor, so it’s slow, but in terms of speed at 533mhz it’s probably not far from the performance of HTC G1, Magic or Hero even though the HTC have faster ARM11 processors.
The main advantage with the ARM laptops is that there are half a dozen or more ARM semiconductor solution providers. TI, Samsung, Broadcom, Qualcomm, Freescale and Nvidia at the least. You can even add a few more processor makers onto that. They are all making ARM Cortex based laptop reference desins and competing on the price.
To me, the only performance ARM needs to reach is to run a web browser like Chrome at full speed, meaning, you click and things happen instantly, and you scroll and it feels just fine, and you open as many tabs as you want and switch tabs and everything seems great. I believe this is possible already with ARM Cortex A8 and will only be improved with A9 in the next few months.
Chrome OS will be released next week, I do believe we will see Smartbooks before Christmas though the mass production may only launch after CES.
Those smartbooks we hear about launching at AT&T are not the core of smartbooks that I think will revolutionize the industry. Cause if you signup for a 24-month data plan, the pricing advantage of ARM is harder to discern for the public.
Though when you talk about the price for a unlocked open laptop, the price for ARM version will be less than half of using the atom of Intel. And battery life truely will be very different, upwards 20 hours on a 3 cell battery with ARM and Pixel Qi while Intel has a hard time reaching 10 hours with Pixel Qi and a 6 cell battery.
Thus, look forward to $100 ARM9/ARM11 entry level laptops, all the way up to $150 to $200 ARM Cortex A8 types. All with 3G support either through dongle or internal, and totally unlocked. All being able to run a full Chrome browser with unlimited tabs and speed of page loading and scrolling, no matter the amount of RAM, and no matter the type of Linux it’s installed on Android, Chrome OS, Ubuntu or some other Linux OS available.
one big question for some will be how good these things will handle something like remote desktop solutions.
and it strikes me that they can be really interesting for some, if they can behave like a mobile phone in terms of always connected. That is, push mail, im status and similar.
Sure, a smartphone can do the same, and one could go with a smartphone and a redfly, either in hardware or software on ones netbook or laptop, rather then a smartbook. But in the end that will be up to the customer.
in the end tho, i suspect whats needed is not new hardware, but new ways to use computing on the go. The laptop or smaller just means we can move our office desk to any flat surface on the planet, but are still doing the same old tasks that be always have done.
it may well be that we need to get rid of the “head down” mode of computing, by having the IO of the computer at the tip of our fingers and nose, rather then in some pocket or bag/backpack where we need to dig it out each time we need to use it.