The more I test Intel Sandy-Bridge based systems the more I am becoming convinced that laptop silicon will eventually extend as a high-dynamic-range platform into to ultra-mobile PCs and tablets. The reason is that the new laptop platforms are using advanced processes and techniques and are extremely efficient at getting things done. ‘Hurry Up Get Idle’ is a simple concept that means if you can get the same job done quickly, you can turn off or idle a pc and thus reduce the power used. The area under the power curve is smaller.
In practice, its difficult to make HUGI work because a lot of the tasks we do are either very short, can only work as fast as a human can input or rely on data coming from other sources. PC’s aren’t very good at idling either but from my recent video editing tests, I can see that there’s at least one scenario where it works very well.
The problem with getting laptop silicon into a handheld product is the thermals. Intel leads the way in this market and their products provide plenty of thermal monitoring and control but it will take a little bit more than what is currently on offer to be able to easily design and produce a 7″, tablet running a laptop-style processor. It’s been done before though. Samsung, experts in electrical engineering, produced a 7″ Tablet running a 1.3Ghz Core Solo but that was at a time when there was no competition from ARM-based devices, $1200 tablets were common and there wasn’t an Atom processor around. The latest tablet example would be the Eee Slate EP121 and for a 1.06KG laptop, the Samsung 900X1A gets close but that’s a little larger and heavier than a handheld device should be.
As silicon processes get better though and thermal control, dedicated silicon and single-chip solutions become more common, you can expect both Intel and AMD to try to offer the ultimate processing power in the handheld space. You can expect these products to have premium prices and to be targeted at niche markets but with Windows 8 as a catalyst and competition increasing from the latest ARM designs, offering these niche product is one way that the X86 chip makers can retain an advantage and one way manufacturers can differentiate their products.
Thanks to Samsung Germany I am now using a rather swish and expensive Series 9 (Model 900X3A) 13″ laptop running the new Sandy Bridge ULV platform. This isn’t a ultra mobile PC of-course but it’s not here for ultra-mobile testing, it’s here for two other things. Firstly I want to use it for the next part of my Ultra Mobile Video Editing series and you’ll see that happen here are UMPCPortal. Secondly I want to start tracking the Ultrabook category very closely as I believe it is not only an important low-power platform for laptops but could, in a few generations, drive sub 1KG Windows devices that will range from sub 1W always-on, to 15W power-houses, exactly what I call High Dynamic Range Computing. They will challenge netbooks for size and power utilisation and include premium features like Wireless-Display, Light-Peek and security subsystems that can be used to stream and store high-value content. There will be a price premium of course but that should reduce to hit the top-end of the netbook range meaning Atom has a positioning job to do. You can follow my work with Ultrabooks over at Ultrabooknews.com
You don’t have to buy a Chromebook to get Chrome OS. Using the Hexxeh open source Chrome OS builds, Lars F. Jørgensen built a Chromebook out of an early netbook, the Aspire One A110. With netbooks in Europe dipping under 200 Euros occasionally now (here’s an offer for a current Acer D255 at Amazon, Germany my Affiliate link) it doesn’t have to be old or expensive either.
I’ve posted the article over at Chromebook News as I feel there’s limited ultra-mobile capability in it but it’s interesting to think of Chromium, the browser, on something like Ubuntu or Meego that may not provide the support-free software layer but would provide all the local facilities needed and, as time goes on, some extended power-saving features too.
Let’s say you need a UMPC. It’s not as uncommon as some people think. It might not be the consumers cup of tea but in industry, mobility counts for a lot. Logistics, amateur pilots, health industry, blue-light industry, traveling geeks and other situations where full capability, compatibility and flexibility in the smallest package is key. The problem is, if you need a ultra mobile PC today, what the hell are you going to buy?
Lets put down a little wish-list for the sake of the argument.
Sub-1KG, Windows 7 support, 5hrs battery life 5-8.9 inch screen, easy conversion to keyboard/screen device. Price under $1000.
The shortlist I would recommend right now would be the following but they are all ‘last-gen’ UMPCs, at least a year old and going out of stock, and probably entering the end-of-life phase.
Fujitsu UH900
Sony Vaio P
Viliv N5
Viliv S7
Viliv X70
Archos 9 (with SSD, 1.2Ghz)
UMID Mbook SE Thx to Gearsguy for the information on the availability and videos. I’ve included one of the videos below.
If you need a keyboard, the Mbook SE, UH900 and N5 are worth a look. The Archos 9 is good value at under 450 Euro right now and the X70 is a great performer. Isn’t it underwhelming that these devices are all over a year old though.
One device I took a second look at was the Toshiba Libretto W100. Originally this device was available for 1100 Euro. Today, it’s under 600 Euro in Europe making it an interesting option because of its CPU – Pentium Dual-Core U5400 with 2x 1.20GHz that comes in at about 130% the processing power of a high-end dual-core Atom part. It also includes 2048MB Ram and a 62GB SSD. This is certainly an ultra mobile workhorse but the design and battery life are going to be issues for some. 3hrs isn’t that exciting. Interestingly this could make a super ultra-mobile video editing platform.
This dearth of options in this space is because of two things. Firstly, Menlow is out and Oaktrail isn’t yet in. There isn’t really another platform to think about right now although I’ve got my eye on AMDs Z-01 We’re going to have to wait for a set devices on Oaktrail for another few months. The other issues is the 10 inch tablet craze. It puts designs at around the 1KG mark and limits usability. The Viliv X70 is one to keep an eye out for but based on the silence from Viliv, I’m guessing it’s not close to being available yet.
Widening your choices
Netbooks, starting at about 1.2KG (2.6lb) and large-format Windows tablets (again 1.2KG when a keyboard is added) along with 5~ and 7~ Android tablets and the iPad2 all need consideration. Even the >4 inch Android phones with the latest CPUs. As Meego filters in, keep an eye on that too as it spans mobile and desktop environments. Finally, Honeycomb and WebOS are operating systems to watch. Personally I have high hopes for Honeycomb as one of the more flexible operating systems to cross-over into a productive and flexible environment and that could happen on either ARM or Intel.
Choosing a platform for 2012
Oaktrail Intel’s Z6xx series. We’ve seen it running Windows, Android and Meego already, it will run Chrome OS and there should be forward compatibility with Windows 8 making it, in my opinion, one of the most interesting ultra-mobile platforms out there right now. Intel builds of Honeycomb and Meego should be able to squeeze more battery life out of it too. There’s a 2X graphics improvement over Menlow (GMA600 vs. GMA500) and even hardware 720p video encoding which could speed up video rendering. At 1.5Ghz, it’s not the most CPU-powerful platform but Intel have already talked about 1.8Ghz versions and I’m sure, if the platform becomes popular, we could see dual-core versions too. Why Oaktrail and not Cedar Trail? Because it’s got power management capabilities that Cedar Trail hasn’t got.
Waiting for Sandy Bridge.
Sandy Bridge in ultra-low-voltage guise is very interesting. I recently tested an AMD-E350 based Lenovo S205. It was good. It’s TDP (CPU+GPU) is 18W and the CPU performance is high-end Atom level. Sandy bridge, on the other had also comes in 17W TDP variants but the CPU performance on these simply blows Atom, E-Series Fusion and even first-gen Core parts out of the water. with around 5x the CPU performance of an Atom CPU along with some good GPU performance. Price is high as we’ve seen with the Samsung Series 9 but that devices comes in at 1.3KG with 6hrs of battery life and serious compute power. It’s a sign that Ultrabooks could push down in to even smaller and lighter designs.
My plan. What’s yours?
Today I sold my last netbook / laptop. Over the last few months I’ve been having a clear-out and now I’m left completely without any sort of mobile productivity device. It’s a nice position to be in but it’s going to be a tough decision. Right now I’m favouring the Samsung TX100 / Gloria / PC7 Slider on Oaktrail because I’m interested in Oaktrail performance and multi-OS scenarios. I’m worried about the CPU performance though. I’m also looking carefully at that Toshiba W100/W105 show above. I think I can run PowerDirector video editing suite on that and get some usable 720p rendering speeds that should be 2x what the Oaktrail platform can produce. Finally, Samsung have another very interesting product in the Series 9 laptop on Core i5 Sandy Bridge. It’s an expensive item but a real mobile workhorse. And why am I looking at all these laptop-style devices? Because after spending 7 months with the Galaxy Tab I’ve found that there are fewer things I need to do on a laptop now and those things generally involve high-productivity working with Video, Images and multiple windows. The 7 inch tablet has filled a great position but along with my new requirement to product 720p videos, has pushed up my requirements for a laptop.
A year ago, Intel teased us with a reference design build on top of a PineTrail CPU. Canoe Lake was a stunning looking netbook but alas, only a reference design. This year though, that could change. Malata are offering it up to their customers.
As you can see though, Malata will build it with the new Cedar Trail platform using the N2600 and N2800 CPU parts. The price, weight, timescale and finer specification details are still unknown but we’ll be tracking this and bringing you the latest soon.
Expect a tracking page up in the product portal later today.
JKKMobile spotted this and has a video, embedded below.
There’s good and bad news to be drawn out of the news that Intel will go with a PowerVR core on Cedar Trail, the next generation of Netbook platform. VR-Zone report that Cedarview (the processing engine in the Cedar Trail platform) will include an SGX545 graphics core and not Intel graphics as had previously been suggested here.
The good news is that it brings the netbook platform in-line with other Atom platforms used for TV, ultra-mobile, embedded and tablet products and simplifies the builds for MeeGo and Android. The bad news is that the drivers for past versions of the SGX core used in Menlow (Z5xx-series CPUs) were never that good. Linux support, in particular, was a big problem.
Intel will, of course, be aware of that and will have to take steps to improve it because if 30 million sales of netbooks have unstable or poorly implemented drivers, it could have a serious impact.
As for the theoretical performance of the SGX545, I can only offer these figures that I dig out of web searches this morning. They aren’t confirmed.
GMA500 (PowerVR SGX535 200Mhz) 13 million triangles per second
GMA600 (PowerVR SGX535 400Mhz) 26 million triangles per second (assumed)
GMAxxx (PowerVR SGX545 400Mhz) 80 million triangles per second (low power version)
GMAxxx (PowerVR SGX545 640Mhz) 128 million triangles per second.
There’s a potential 10x improvement over GMA500 there but these figures are fairly meaningless as memory bus, cpu processing power and other implementation issues (including driver software) will affect the true performance. The diagram shown on VR_Zone though shows a target of 2x Pinetrail. While that’s probably not going to beat the 3D performance of AMDs Zacate platforms, it’s a welcome boost that will drive non-gaming devices well and, we assume, keep the power requirements down. For non-gamers, the balance of CPU, GPU, media and power drain is likely to be good.
My outstanding question:Will Intel enable ‘shutdown’ idle (sub 200mw system drain) for screen-off scenarios on Cedar Trail like it is trying to do on Oaktrail, Moorestown and Medfield? This is something they may work with Microsoft on for Windows 8
The news has not been officially confirmed by Intel but that might happen at Computex later this month. If not, wait for hands-on at IDF 2011 in September where we expect final launch and product availability to be announced.
UMPCs, we’re not shy to say, are still important. They’re becoming more important too as people realise that there is a life for devices between 4 and 10 inches. UMPCs offer something that you can rely on to give you a high quality work and web experiences with the minimum of brick walls and maximum compatibility and flexibility.
Unfortunately for pro-mobile fans, the number of true UMPCs has dropped to near-zero with only large-screen windows tablets getting any sort of attention. It’s the Android and IOS tablets that are getting the lions share of attention and development funds though and we are now in a position where we need to keep an eye on the desktop and mobile worlds for solutions that could satisfy our needs.
I’ve done extensive testing and research on the smartbook market over that last 18 months with devices like the Compaq Airlife 100 and Toshiba AC100 but I haven’t found anything that I could call pro-quality. While the industrial designs and computing hardware reach acceptable levels, the immature operating systems and applications truly limit capability and introduce inefficiencies and risk.
Honeycomb, the tablet-oriented version of the Android system is the next-in-line for major testing and the Asus Eee Pad Transformer is the ideal product to test it with.
Image via Laptopmag
Although I haven’t yet been able to get hold of a Transformer as my Amazon UK order was delayed to June, I have been carefully reading between the lines of the many reviews that are out there. I’m trying to get a feel for web experience, applications, keyboard usage, connectivity and battery life along with an idea about how those ‘HD’ apps are coming on.
So far, I am not seeing an acceptable solution being presented.
While a lot of reviewers seem impressed with the price and design of the tablet, keyboard and docking feature, the battery life and general usability of Honeycomb, there seem to be issues with Web (once again, not the full desktop internet experience) performance and applications. It’s the latter that concerns me the most as it will take the longest to fix and relies on a healthy Honeycomb ecosystem to even seed. I’m also not so impressed with the price. It may be cheaper than other tablet options but it’s still an inflated price for what you’re getting. Finally, at 1.3KG, it weighs 130gm more than the Aspire one 522 that i’m using right now that has nearly double the CPU processing power and 3D graphics that just don’t compare along with an OS developed for professional use and a huge range of professional software available.
Keyboard – Good ID
In general, the keyboard and mousepad are getting good reports. A good level of mousepad/OS integration ensures that the unit is not relying on touch input when docked. The battery in the keyboard unit is used to recharge the tablet battery (a very inefficient, but cheap, way of using an extended battery) and the USB ports are of limited use. Mass storage and external mice are supported but don’t exptect to be able to plug in a webcam, usb headset or many other common USB accessories that ‘just work’ under Windows.
Battery Life – So-So.
In total, the Asus Eee Pad Transformer returns around 9hrs in tablet mode. Some reviews are reporting down to 6hrs. Usage in docked mode is going to depend on whether you dock the device when the tablet is fully charged as the battery in the docking station will try to re-charge the battery in the tablet while also powering the system. The battery to battery re-charge process is going to cost 20-30% of energy. From results I’ve seen the maximum you can expect from the docked unit is about 16hrs which, with a 48wh battery, shows that ARM-based devices ultimately rely on screen, wifi and engineering to get the best out of them, just like X86-based devices. I haven’t seen any screen-off, wifi on figures yet (always-on mode) but am hoping for at least 4 days. If the Transformer is very well engineered, that figure should go up to the 7-10 days range. ElectricPig indicates that the Transformer loses about 10% of battery with Wifi on in a screen-off, syncing on scenario. Assuming that’s about 10hrs of idle, it indicates a reasonable always-on, Wi-Fi connected capability of about 4 days (tablet only) or about 6 days (tablet and dock.) That beats the Toshiba AC100 my a big margin.
Web Experience – Not FIE
Tabbed browsing is nice, but many reviewers are focusing on this , and the Flash 10.2 experience as a big step forward. For productivity users, these are entry-level features. While thre aren’t too many details in the reviews I’ve seen so far there are indications that the quality of the browsing experience is compromised. I’m particularly interested to hear how Google Docs works with the keyboard. AnandTech reports ”occasional issues’ that include a poorly formated Reddit front page.
Applications – Back to the Smartphone
As I mentioned above, this is the major issue for Android Smartbooks and it’s not one that manufacturers can solve. It will only resolve itself when Honeycomb gets traction, proves numbers and drives teams to develop quality software over a period of months. At the current rate of change in software, we’don’t expect Honeycomb to offer anything like a serious 3rd party productivity application range until at least late in 2011.
Other issues to consider.
No 3G / 4G
No VGA output (for projectors, monitor compatibility)
Video playback not smooth (as reported by a number of website reviews.)
Some reviews are reporting speed slow-down after a couple of hours of use.
Sidenote: The Anandtech review is an interesting read that explores the idea of continuous client and smartphone-based modular solutions.
Summary.
For smartphone and tablet users, the Asus Eee Pad Transformer must appear quite exciting. It did for me too but I knew there would be issues. The issues of application availability will haunt every attempt at an Android productivity device for at least the next 6 months, probably longer, but if manufacturers fail to capitalise on battery life possibilities and if Google can’t update Honeycomb to finally support a Chrome-like internet experience, there will be limited reason to accept these attempts into the realms of profesional environments and little enthusiasm for creating rich, high quality, professional applications. The few points mentioned above are obvious failings but there are other hidden items that will rise to the surface as people truly start to try these devices out for productivity. Think about multi-user capability for example. Disk encryption, memory upgrade, VGA ports and 3G support are just a few more to consider.
I’m disappointed in the battery life reports, video playback issues, weight and web experience that I’m reading about so far. So much so that i”ve decided to skip the Eee Pad Transformer. I will try to get hold of a review sample but it won’t be something I’ll take on board into my work life. Like the Airlife, AC100, Atrix and other attempts, there’s so much missing, so many brick-walls, so few applications that I don’t expect any major progress until later in the year. The Transformer proves the point that Honeycomb can move in the right direction and as a tablet device, it may well be good value but for pro-mobile users, I’m not recommending it.
Caveat -These thoughts based on my experience with smartbooks, my own hands-on with the Transformer, feedback from contacts and information available in the public domain. Find full reviews of the Asus Eee Pad Transformer in our product page. Reader, owner feedback is encouraged below.
Technical sessions at Intel’s IDF in Beijing have all finished now so it’s time to go through some of the presentation material, the press releases and interview information from various sources to put together a summary of what Intel are planning with Cedar Trail in the netbook market. some details are still misssing but at this stage we can put together a fairly complete picture.
Intel’s netbook strategy comprises two platforms. The first, and the one that appears on most netbooks, is the Pinetrail platform. That is due for an update later this year and the new platform will be known as Cedar Trail, the one we’re discussing here.
The second platform is a more specialist, low end [performance] platform that grew out of the Menlow ultra mobile PC platform. Oaktrail, which uses the Z6xx Lincroft cpu is shipping now and offers a lower TDP with refined graphics and enhanced power-saving features. Oaktrail is for embedded, tablet and thin, light and rugged netbook solutions. Information on Oaktrail can be found here.
Netbook market predictions.
While excitement in the netbook sector has cooled off somewhat and most commentators agree that some parts of the market are seeing competition from tablets, the sales numbers remain significant.
Note that this is the total addressable market. AMD Fusion, ARM and VIA-based devices will be competing for this market too. For Intel to compete it needs to address the top, bottom and specialist segments of the market.
