I’ve already posted about this, twice. Samsung have some great engineers.

I had the chance to look at an NC10 (#1 device on the portal right now) at the netbook meeting in Cologne on Saturday and the first thing I did was to measure the minimum drain of the device. Within minutes, I was seeing a fantastic sub-5W minimum drain.

This is lower than on most UMPCs and is a good indication that the electronics on the motherboard, the power components and the screen are well engineered. There’s no point putting in a low-power CPU if you’re going to lose the advantage it with crappy engineering. Take the Kohjinsha SC3. It uses the latest, lower clock, lower power CPU and chipset and yet even after hours of optimising, the base drain on it is no less than 7.5W. That’s 50% more than the, much bigger NC10 that doesn’t even use the low power CPU and chipset.

Battery life outweighs most requirements for me so considering that everything else on the device is up to scratch, I’m seriously thinking of getting a black NC10 as my only laptop and selling the M912 and Medion Akoya. They’re both excelent devices but neither of them are as well engineered as the NC10

Samsung NC10 links and specs in the product pages.

One of the distributors of the Wibrain i1 has got hold of a Wibrain i1, snapped a few pics and made some comments in the forum here. If you’re looking the the ultimate Windows XP mobility device with the best battery life and best connectivity for, potentially, a very interesting price, keep reading.

The 1.3Ghz Atom-based device is based on the old Wibrain B1 design which was, quite frankly, an ugly brick to most people. There are some styling improvements which help it a little but don’t expect young geeks to be fainting at your feet when you whip it out. However, this is an action UMPC, not a flowery style-icon. It has the same easy-to-learn control layout (the touchpad is the best I’ve ever tried on a UMPC for example) and the same extremely high quality and brightness screen in a package that weighs 500gm. That’s 1.1lbs. It runs XP, has a 60GB hard drive, new SD card slot and re-positioned USB port. The real kicker is three-pronged though. 6hrs+ battery life on the standard 30wh battery, a 3G module and what looks like it could be an amazingly good price for such a setup.

Wibrain claim 7hrs on the standard battery. I’ve seen 6+ on the battery meter in hands-on testing and ‘Digital’, is now reporting 6.5hrs with wifi or 3G on. In marketing terms, this is an all-day UMPC in 500gms.

Final availability is still not 100% clear but early pricing indications at Mobilx look positive. 467 Euros ($600) pre tax without 3G and 532 Euros ($712) pre-tax with the 3G module. We’ll keep you updated on availability when we hear anything.

More links, news, details in the Wibrain i1 information page.

Remember my in-tent Samsung Q1 Ultra test? 9hrs battery life on the Samsung Q1 with 25% backlight? An average drain of 6.9 watts. It was impressive and since I blew up the Q1 Ultra doing some solar-charging tests, I’ve been missing those 5, 6 and 7hr figures. I was just getting over the trauma until I read this article about another Samsung Mobile product with great battery life.

Notebook Magazine have just done a full set of battery life tests on the NC10 Netbook and it looks like Samsung have once again worked their magic. The tests included continuous Wifi-on web browsing in battery saving mode which is a very reasonable test to be doing. Far better than the misleading 2001 Jeita test and much more ‘real-life’ than the Battery Eater Pro ‘turn everything and and run it at full-steam’ approach. The NC10 returned…

• 12% screen brightness: 7hrs 34mins representing an average drain of about 7.6w which is just a little bit more than what I saw on the Q1 Ultra.
• 50% screen brightness, the duration dropped to 6hrs 30mins which is an average 8.7W.
• 100% screen brightness, the test result reduced to 4hrs 38 minutes which is 12.3 watts drain.

More after the break…

(more…)

I recall seeing this solution a while back on JKKMobile but Paul from MoDaCo has now got one for testing. Its a 47Wh battery pack for the HTC Shift from Mugen which should allow about 5hrs online time thus solving one of the HTC Shift’s biggest problems. It’s a tidy solution but at nearly $240, it’s a huge amount to pay for 5 hrs of computing time.

Paul is doing further testing and will report back on performance soon.

Kevin C. Tofel over at JKotR has spent some time benchmarking the 6-cell extended battery for the MSI Wind. I have to say, I’m really impressed. Using BatteryEater, Kevin was getting 3 hours and 39 minutes running the minimum runtime test, which uses the CPU at 100% capacity until the battery dies. Running the test a 50% screen brightness, Wi-Fi on, and no Windows power management, it is easy to how a regular work load could give you nearly 5 hours of runtime. Kevin also ran BatteryEater after toggling the Wind’s ‘Turbo’ button, which cuts the CPU speed in half. Running at 800MHz, the Wind went for an impressive 4 hours and 59 minutes under full CPU strain. Kevin says that with regular usage and power management, you could hit 7 hours with the extended battery. Head over to JKotR to see the full article including a comparison to the Asus Eee 1000H.

I’m in the ‘BUX (as my fellow US-based bloggers say!) with the SC3 and having a fun, productive time. Watching that battery level of course but really enjoying the form factor. To most people round here I guess it looks like ‘one of those small notebooks’ but they are so so wrong!

Things you can’t do with a netbook..but can with an SC3.

• Thumb the keyboard. Yup, it feels comfortable for one-lines, IM, passwords and URLs.
• Hold it in one hand and drive it via the touch screen while leaning  back.
• Leaning back with a coffee in one hand and the SC3 converted to tablet mode reading feeds in the other.
• Slipping it into my ‘man purse.’ [Yes, a debateable advantage ]
• Bluetooth teathering to my phone (OK, some netbooks do actually have Bluetooth)
• Feeling really happy that you have one of the smallest and most adaptable PC formats ever.

Problems you must cope with…

• Battery life
• Keyboard not as good as a netbook but im happy to sacrifice a handful of words per minute.
• Vista keeps slowing me down with its disk activity and inability to be slim and fast. I checked out an XP-based 1.33Ghz Silverthorne-based device this morning. It was so much better.

I have decided to give the SC3 another chance. My Tekkeon and Samsung Q1U died last week so rather than shell out $1400 for my ideal Q1 Ultra, i’m going to get a slim, 30hw external battery pack that will fit in my gadget bag and provide some badly needed energy. When I get the chance, i’ll try out the Kohjinsha extended battery. It should save me $1000 while I wait for a new device. In the meantime i’ll have to put up with the rollercoaster of emotions that the SC3 gives me. 2hrs mobile happiness followed by massive frustration that device is unuseable without mains power! A bit like the HTC Shift but with a much, much better screen.

[Note about that rollewrcoaster: I was tethering to my N82 rather than using WiFi to save battery power and the battery on my phone died before I had a chance to post this. Battery life, battery life, battery life.  Grrr!]

Last week I posted some tips on how to squeeze 3 hrs browsing time out of the Kohjinsha SC3. The method was a bit of a trick as it utilised a mobile phone data connection over Bluetooth but it was the only way to achieve 3 hours on the 20wh battery. It equates to an average 7W drain which is, even in UMPC terms, very efficient but it’s not as good as I hoped from a Z-series Atom-based system.

The reason why it’s not as good as I expected (and the reason why i’m not using the SC3 to write this post in my tent while the rain beats down outside) is that the Samsung Q1 Ultra (with SSD mod) is just as efficient with the previous generation Intel platform. As I write this post, perfom (try it, start->run->perfmon. Its fun to watch the battery drain) is telling me that the average drain is 6.9 watts. Meebo is running in the background, Friendfeed is updating, CPU utilisation is averaging 20%, brightness is set at 25% and I’ve got a LED lamp attached to the USB port to give me some light. With the standard battery, thats over 4 hours of online time. I’ve got the extended battery here which is showing 5.5 hrs @ 61%  battery. (9hrs total)

With a refresh to a 1.33Ghz Atom processor and some tweaks to the motherboard, I’m sure Samsung could shave 20% off that drain figure and produce a 5-6 hr, 600gm device. They’ve proved themselves as one of the best at electronic design with their Q1 products and it excites me to think about what they might come up with next - as long as it has a lighted keyboard!

I’m going camping over the next four days. Here’s the kit that’s going to keep me productive.

[The Wife and little-one have planned a last-minute camping trip together and I was due to stay home and work but I've decided to go along and make a working holiday of it. Dads - this is the beauty of Ultra Mobile PC's!!!]

That’s a Samsung Q1 Ultra being charged from a Tablet Kiosk MP3400 which will be charged using a Sunlinq 25W solar panel. Also included are the Nokia N82 and a battery charging set. More details here.

Check out Solar-UMPC for some more mobile and solar-computing posts over the next few days.

A clear message to UMPC OEM’s

If it wasn’t clear before, it is now. [If the results are not showing, make your vote and you'll see them.]

While some people will accept 3-hours battery life for an Ultra Mobile PC, 85% of people want 4 or more hours. 50% of people here see 5 hours as minimum battery life for a UMPC.

As you might have read, I have been very disappointed with the battery life figures on the Kohjinsha SC3. I was expecting a lot lot more from an Ultra Mobile PC built on Intel’s latest battery-optimised Menlow platform but it turns out that under normal use, this device is just as bad as many devices based on the previous platform. Over the last 24 hours I’ve spent a lot of time trying to analyse why and have finally come up with the reason. Invetec, the OED for this device, have cut corners. While the background drain (on the motherboard and power board) is better than on previous Kohjinsha’s, it’s no better than the HTC Shift, Q1 Ultra. All these devices can match the SC3 for background efficiency. About 4W minimum or 5-hours, is what you can expect with everything turned off but the processor.

OK, 4W is good but once the screen is on, boom! Up goes the battery drain by between 80 and 150%. This is exceptionally high drain for what appears to be a LED-backlit screen. Thank goodness that using the screen in a bright room requires only 3/8th brightness but it’s still about 3W of drain which is much higher than it should be.

The second shock comes when you turn the VIA, yes VIA!,  USB Wifi module on. Connecting to a hotspot will push the drain up by over 3W. It will settle back to 2W after connection but this is, quite frankly, pathetic. A wifi module that takes more power than the processor is a joke. There are far more efficient Wifi modules out there. Did Inventec think they were designing a cheap netbook?

So from a background drain of 5W, if you want to surf the net in a bright room over Wifi, the battery life drops to a depressing 2hrs. This is an average drain of 10W which is no better than the devices mentioned above.

Breakdown, tips and summary follow…

(more…)

Lets start this article with a poll.

If there’s one thing that always annoys me in the UMPC category, it’s a wasted chance. Be it poor design, poor software or poor hardware. If there’s one thing that annoys me more than that its when a company makes the same mistake twice. Even worse than that is when the device in question is high-quality in every other respect. Because of this I find the SC3 the most frustrating device I’ve ever owned. I want so much to love it and take it into my daily life but….

(more…)

Author: Hanno Zulla
Original article appears on Hanno’s blog. Reprinted in full with permission.

When it comes to portable electronic gadgets, there are three major annoyances.

The three hour limit must fall

Photo by AndyArmstrong

There is an unwritten rule when designing portable computers:

The battery will last three hours.

Once technological advances allow the next generation to run longer - be it thanks to more efficient hardware or more powerful battery technology - the manufacturers decide to shrink the battery, capping the device back to the three hour limit.

Three is a nice psychological figure. “Lasts three hours? - not too short!” “Less than three kilogramms? - not too heavy!” After all these years, today’s 3 kg laptops usually still run for 3 hours or less.

This must end.

Three hours is not enough for a true mobile device (especially since the advertised three hours of battery time usually result to less than two in real use).

Read on for more….

(more…)

At CeBIT 2007, Intel revealed their ‘Menlow’ plans and told a room full of reporters that we’d be seeing sub 4W PCs in 2008. In March 2008 I asked Intel if they were on target and they told me that I’d be surprised. I am!

I’m lucky enough to be testing a Menlow-based Gigabyte M528 (sample) this weekend and this evening I’ve been doing a battery life test. I’m listening to MP3’s streaming over Wifi through a java applet on the Firefox 3-based browser with the CPU averaging about 30% utilisation (yes, this java app isn’t the most efficient at playing MP3’s!) It’s been going for over 2.5hours on its 10wh battery. (backlight off)

For those that don’t know, the Gigabyte M528 is one of the first MIDs and it’s based on Intel’s Menlow platform; A PC platform. This baby can run Windows XP!

The test is still going. Check back to see how long the test goes for….

Updates:

I’ve just passed the 3hrs mark. I hope the battery dies soon as its 00:20 here and I need to go to bed!!

It’s 12:30 now. 3hrs 10 mins…still going…

I estimate that with no backlight, the device is running about 0.5W less than with the screen backlight on at 50%. When the test finishes I can estimate a backlight-on time.

3hrs 20 mins. That’s an average drain of 3W. Its getting into smartphone territory at this point.

3hrs 30 mins….

Battery light flashing at 3hrs 35 mins… not long to go now.

And there she goes. Test completed at 3hrs 35 mins.

Average battery drain (assuming the battery really is empty now.) 2.8W

I estimate that with the screen on, it would be about 3.3W which would give a run time of 3hrs.

Well done Intel! I’m off to bed now.

This tickled me. A post on Friendfeed a short while ago…

“As the iPhone begins to resemble and be used like a computer, we will have to contend with challenges like crashing and battery life.” [Steve Rubel]

He’s right. Many people still assume that smartphones have excellent battery life compared to UMPCs. The reality is slightly less exciting if you use both devices for mobile Internet activities on a regular basis.

Apart from having a little giggle, the comment also reminded me to check up on the reported iPhone 3G battery life and compare it with my figures. I came up with an interesting article from AnandTech who have completed a browsing test which resulted in a dead iPhone in just over three hours. That’s much less than the 5-hrs quoted by Apple. It’s not even the most power-hungry test you could do on an iPhone. Try doing some Voip, live tracking or even IM with some music playing in the background and you’ll have a dead iPhone in under 2 hrs. Well, you would if the Iphone would let you run 3rd party background tasks!

3G radios, GPUs, displays are all going to take a lot of power whether they’re on a UMPC or a smartphone. Add 3rd-party software into the mix and you find that if you want the FIE on a regular basis, there’s no substitute for a fat battery!  As I mentioned in a previous article, 10wh is about your minimum battery size for any serious mobile Internet device in 2008 and 2009. The battery in the iPhone is only 5wh (possibly only 4wh according to this article.) 

In perspective, the iPhone 3G is still an impressively efficient device. 1.5W maximum power drain makes it one of the most energy efficient MIDs around and it’s still leading the way in the consumer-oriented market. In 2007 we were seeing best-of-breed UMPCs taking 9W to do similar 3G browsing tasks. Later in 2008, the differential will shrink to just 2 or 3watts and in 2010, the differential will be close to zero but in all cases, you still need that big battery if you need the FIE for long periods between charges. For now, the multi-device strategy still seems to be the best for you pro-mobile users.

[If you spot any other iPhone 3G battery life tests that were done under 'MID' conditions. Please drop me a comment. I'm interested. Thanks.]

Energy efficiency is, without a doubt, the most important design factor for any pocketable Ultra Mobile device. On Friday this week, there are two very important launch events taking place that will highlight the advances made in energy efficient computing over the last year. The two devices will also show how us close Intel and ARM partners really are in the race to provide handheld, Internet-focused computers. MIDs.

On one side of the fence, in PC land, Sharp and Willcom will release a pocketable slider-keyboard mobility-focused device. The Sharp Willcom D4 will be running Windows Vista on the Intel Silverthorne/Poulsbo platform also known as Menlow and is likely to be one of the most energy-efficient 3G-enabled PC’s so far.

On the other side of the fence, in smartphone land, Apple will release a smaller, 3G-capable mobility-focused device running a highly optimised desktop operating system. The iPhone 3G will be running ARM-based processors and is likely to be one of the most energy-efficient 3G smartphones so far.

The Willcom D4 is said to have (independent tests) a 3G-active time (Internet activity) of around 1.5 hours and it runs a 7wh battery.

The iPhone 3G is said to have (Apple figures) a 3G-active time (Internet activity) of around 5 hours and it is likely to run (my estimates) a battery of about 5.5wh.

3G-active Internet time is a worst-case measure of device efficiency so, based on these early figures, it looks like the Menlow-based device will use about 4.5W and an iPhone will use about 1W.  More after the pic…

(more…)