Microsoft Surface tablets

[wco81]

Weight is definitely an issue for extended reading, especially lying down. If you sit up a bit instead of lying flat on your back or on your side, it's better but not everyone has that kind of a setup.

I like the iPad but I don't take it on the go that much. Take it to the office, that's about it. I can definitely see using a smaller tablet for reading, GPS turn-by-turn (better chances of finding a mounting solution) and probably take it out and about.

I bought some tourist guide ebooks for a recent trip but only took the iPad out of the hotel a couple of times at most (had room in my messenger bag but when you're walking miles and miles, you want to minimize the weight you carry, though I was lugging pounds of photo gear and other stuff already).

I would imagine someone will put out smaller form factor W8 tablets? Maybe not x86 ones because of power requirements but that may be an area where the W8 ARM tablets find more acceptance? Because a lot of people seem to be of the opinion that the x86 tablet makes the RT one superfluous, though just on the expected price differential alone, you'd think RT would have a good chance for sales relative to the x86 one.

 

[Andrew Lauritzen]

Because a lot of people seem to be of the opinion that the x86 tablet makes the RT one superfluous, though just on the expected price differential alone, you'd think RT would have a good chance for sales relative to the x86 one.

Well from a capabilities point of view it is a strict superset of the RT version. The only advantages of the RT AFAIK are being thinner, lighter and cheaper. Obviously these are relevant advantages to a lot of people and I fully expect it to sell more than the Pro simply due to price, but on Beyond3D I'm not surprised to see the audience more interested in the technical capabilities than the price.

I think the RT will find stiffer competition in the market though, as it really is just a direct ipad competitor with only a few small differentiating features. The Pro is its own thing really, with only the Samsung Series 7 Slate having done something similar and no real non-windows competition. (Closest would be putting Linux on an ASUS Transformer or something, but I don't imagine a ton of folks are doing that.)

 

[Silent_Buddha]

Also: With a reflective screen, the brightness of the screen is always relative to ambient light: Never too bright, and never too dark.

Cheers

This is a big one in my experience. Especially when lighting conditions change.

With e-Ink you never have to adjust brightness and/or gamma if your room lighting changes. Nor do you have to attempt to keep your room lighting at some constant in order to maintain a properly lit screen with respect to ambient lighting. Is it even possible to adjust gamma on most tablets? I'm sure you can change brightness, but other than Windows based slates, I haven't got a clue if you can change gamma on an iPad or Android slate.

Regards,
SB

 

[Xmas]

To be fair, when you simply compare the use cases enabled by e-ink, gamma is hardly an issue.

Fringing...put an opaque object in front of a light source and the light at the edges diffracts around the corner (the edge is like a line source with light moving away at all angles including in the direction of the opacity).

Persistence is just the reciprocal of flicker. e-ink has no flicker at all.

Can't say that I can associate fringing with any specific visual impression when looking at an LCD.

LCD backlights are usually pulsed at a sufficiently high frequency that flicker isn't an issue.

 

[Simon F]

Do adults use night lights?

I don't know about adults, but children probably should not be given night lights.

 

[Mintmaster]

Well from a capabilities point of view it is a strict superset of the RT version. The only advantages of the RT AFAIK are being thinner, lighter and cheaper. Obviously these are relevant advantages to a lot of people and I fully expect it to sell more than the Pro simply due to price, but on Beyond3D I'm not surprised to see the audience more interested in the technical capabilities than the price.

Yeah, but those advantages are very short lived. Medfield is getting competitive power consumption in a smartphone, and Clovertrail will do the same for x86 tablets. That removes the thinner/lighter/longer-lasting advantage. We've already seen netbooks hit extremely low price points, and tablets with the same hardware will have an even lower cost floor. So cost won't be an advantage either.

Sure, the Surface RT will be significantly cheaper than the Surface Pro, but they're just examples to help lead the market in the right direction. There's no future for Windows RT. Even Android has rather limited adoption in tablets, so it's unlikely to survive the incoming onslaught of Clovertrail tablets beyond a couple years. ChromeOS and Linux tried to lower costs through cheaper hardware and a lack of licensing fees for netbooks (the precursors to tablets as internet consumption devices) but they still got clobbered by Wintel.

So really, I don't expect the RT to sell better. The thing is that W8 tablets aren't just attacking the tablet market, they're attacking the notebook market as well.

 

[RudeCurve]

So really, I don't expect the RT to sell better. The thing is that W8 tablets aren't just attacking the tablet market, they're attacking the notebook market as well.

Anything with a tiny 10" screen is not a notebook...that's what you call a netbook. I doubt most notebook buyers will want to go for a $1000 W8Pro tablet instead of a proper notebook with large screen. The rest of the population will just buy the cheaper "proper" W8 RT tablet or iPad. W8Pro tablet = netbook with touchscreen and detachable keyboard...been there done that about 10 years ago...

 

[Laurent06]

Yeah, but those advantages are very short lived. Medfield is getting competitive power consumption in a smartphone, and Clovertrail will do the same for x86 tablets. That removes the thinner/lighter/longer-lasting advantage. We've already seen netbooks hit extremely low price points, and tablets with the same hardware will have an even lower cost floor. So cost won't be an advantage either.

Clovertrail will certainly disappoint people who want to run x86 software. It won't very likely be faster than the current high end Cortex-A9 SoC. Even netbooks were considered too slow by many people and in the end Clovertrail is again the same core that was put on the market years ago.

 

[french toast]

Clovertrail will certainly disappoint people who want to run x86 software. It won't very likely be faster than the current high end Cortex-A9 SoC. Even netbooks were considered too slow by many people and in the end Clovertrail is again the same core that was put on the market years ago.

Yea having experienced running (or trying to) x86 software on an atom..I have to say it is an experience I wouldn't want to repeat.

Medfield wasn't really that competitive battery wise even if you look at the pro Intel anandtech.
Too be fair to atom it did a lot better than anybody thought it would...certainly if I was looking for a budget smartphone I would pick up the orange san Diego in a heartbeat...fantastic phone for its price..astonishing even..

Beginning to go with public opinion on this...if you want a w8 tablet...go with x86...probably haswell being the sweet spot.

 

[Silent_Buddha]

Anything with a tiny 10" screen is not a notebook...that's what you call a netbook. I doubt most notebook buyers will want to go for a $1000 W8Pro tablet instead of a proper notebook with large screen. The rest of the population will just buy the cheaper "proper" W8 RT tablet or iPad. W8Pro tablet = netbook with touchscreen and detachable keyboard...been there done that about 10 years ago...

Erm, really? Would you be so kind as to point me to any one of those that were even remotely affordable? I've been paying close attention to that market for the past 10 years and I must have missed all those relatively affordable (~1k USD) Windows slates. Most of the XP versions ran around 3K+ USD with some hitting 10k+ USD. Vista brought some slates into the 1.5-2k USD region. Win7 dropped them to sub 1k USD with the drawback of using Atom processors (so netbook in a slat) with most of the affordable ~1k USD i3/i5 slates not coming until later.

True there were convertable tablets which were basically just Notebook computers with a swivel screen, but I can only think of one of those off the top of my head that had a detachable keyboard and that particular model was over 2k USD.

Win8 does many things to make a Windows slate more accessible to the public. Rules for ergonomics for Metro apps means a high likelyhood of applications that are touch friendly. A major drawback when using the vast majority of Windows applications on a slate in the past 10 years. While Vista and Windows 7 UIs were quite useable on a touchscreen for the most part, the same couldn't be said of the applications you could run.

So, WinRT is basically only for casual slate users. People that are currently using iPad and Android slates for the most part. Will it last? Depends on how performant ARM actually is compared to Intel and AMDs x86 parts as well as how robust the Metro line of applications are.

Windows x86 slates have ALL the benefits of WinRT with the following exceptions. First gen devices will still be slightly thicker and slightly heavier. That will likely be addressed as Intel moves ultra power efficient CPU's onto its modern process fabs along with increased R&D into ULV processors. On the other hand, they also bring the capability to run anything on the slate that you could on a desktop or notebook.

Sure, the Microsoft Surface might be too small to be a proper notebook replacement for some people. But that's fine. There will be OEMs offering 11-14" slates at some point. Asus already makes a 12.1" Win7 slate for example. I wouldn't be surprised if at some point one of the OEMs decides to make a 15"+ slate. Hell, there's no reason some company couldn't make a 30" Windows slate? Why would they? I have no idea.

Considering you can use any off the shelf keyboard and mouse with a slate there's also the potential for desktop replacement slates capable of gaming with similar battery life to desktop replacement "laptops" capable of gaming.

Basically just about any form factor or solution you can think of will be able to be addressed by Win8 x86 as well as WinRT. Desktop, laptop, notebook, netbook, convertable tablet, slate, slate w/detachable keyboard, whatever. With metro apps being usable across the entire range, and desktop apps being available on any x86 device.

Regards,
SB

 

[Mintmaster]

Anything with a tiny 10" screen is not a notebook...that's what you call a netbook. I doubt most notebook buyers will want to go for a $1000 W8Pro tablet instead of a proper notebook with large screen. The rest of the population will just buy the cheaper "proper" W8 RT tablet or iPad. W8Pro tablet = netbook with touchscreen and detachable keyboard...been there done that about 10 years ago...

Anyone who wants portability won't mind, especially with 1080p resolution. I'm not saying it can replace all notebook needs, but the MacBook Air is almost 30% of all of Apple's notebook sales, and 10.6" is not much smaller. There's at least a 20M/yr chunk of notebook sales where the purchaser would take a hard look at an x86 hybrid instead plus another 30M/yr netbook sales.

This is pretty big considering that non-iPad tablet sales are probably going to be under 30M this year with MS only getting a small piece of that. So I don't think the RT will sell as well as the Pro.

Yea having experienced running (or trying to) x86 software on an atom. I have to say it is an experience I wouldn't want to repeat.

A lot of the poor experiences people had with the netbook was due to the single core Atom and slow HDDs (often 4200 RPM or really bad stuttering SSDs). Give Clover Trail a modern 64GB SSD and it'll be at least as fast as any ARM tablet at tablet tasks, and that's all it needs. The single core Medfield is already handling browsing and Javascript faster than the quad-core SIII. There's a million windows apps out there that don't need Core i3 speed.

There really isn't anywhere for ARM to go aside from the Apple faithful. Displays now take up such a large chunk of power consumption that even if ARM SoC makers had access to Intel's fab technology (and they don't), it still would be unlikely to make a big enough difference in battery life (or weight) to be a selling point.

 

[RudeCurve]

I think the 10" W8Pro Surface will be a flop/niche product. It doesn't know what it wants to be so it tries to be everything. As a pure tablet it will be expensive for what most people use tablets for...consumption. As a notebook the tiny 10" screen is basically useless for "real work" regardless of 1080p resolution. As a desktop replacement you could use an Ultrabook for the same purpose and use a separate wireless keyboard and monitor. I see very little reason why it would suddenly become the mainstream form factor. I think if the screen was slightly larger around 13" it would sell much better.

 

[eastmen]

I think the 10" W8Pro Surface will be a flop/niche product. It doesn't know what it wants to be so it tries to be everything. As a pure tablet it will be expensive for what most people use tablets for...consumption. As a notebook the tiny 10" screen is basically useless for "real work" regardless of 1080p resolution. As a desktop replacement you could use an Ultrabook for the same purpose and use a separate wireless keyboard and monitor. I see very little reason why it would suddenly become the mainstream form factor. I think if the screen was slightly larger around 13" it would sell much better.

well


1) Schools . The pro verison would work great with the pen input for smart boards . Teachers can go around take a snap shot using the camera of a students work and then correct it up on the smart board .

2) Busniess , its tiny and you'd be able to take it with you anywhere you go , with a core i5 or i7 it will be fast enough for 90% of the work a busniess man/woman would , its lighter than a ultrabook.

So why carry an ultra book around ? When your at the office you can have a nice 27 inch screen , keyborad and mouse . When your at home you can have the same thing. While on the go you have a nice small tablet with a keyboard and track pad , a pen input or a touch.


A 13 inch screen would make a very large tablet , it could work and the iconia w700 is really interesting to me as well as the surface , but i think its to big to be portable, that could be a good home tablet.

 

[wco81]

Schools are buying $1000 computers?

 

[eastmen]

Schools are buying $1000 computers?

Yea , do you not see all the schools buying ipads and macs ? A surface would be a much better value for the teachers. We'd keep going for $600 dells with ssds for the kids and well we don't update the imacs or mac books anymore , they don't hold up to the teenagers well at all.

 

[RudeCurve]

So why carry an ultra book around ?

Because I can do real work with it anywhere because it has a large enough screen. Why would I buy a "portable" $1000 desktop replacement with a itty bitty 10" screen and pretend I could do real work while on the go? Why have all that horsepower "on the go" when the screen limits its use for real work?

When your at the office you can have a nice 27 inch screen , keyborad and mouse . When your at home you can have the same thing. While on the go you have a nice small tablet with a keyboard and track pad , a pen input or a touch.

Why would I replace my home desktop with this device? Why should I get this instead of an Ultrabook that has a bigger screen. Why should I get this instead of the cheaper RT version if both can't do real work because both have little 10" screens which means they're only good for consumption?

A 13 inch screen would make a very large tablet , it could work and the iconia w700 is really interesting to me as well as the surface , but i think its to big to be portable, that could be a good home tablet.

It wouldn't be very large especially if the bezel was thin...besides neither one is pocketable so why does it matter? You'd still have to carry a slate around.

 

[Mintmaster]

Why have all that horsepower "on the go" when the screen limits its use for real work?

I don't know where you're getting this nonsense from. How does a 10.6" with 1080p limit its use for real work so much more than a 768p 13"? IBM/Lenovo, HP, and Dell have sold tens of millions of notebooks with 11.6", and 12.1" screens to business users with lower resolution. Millions of netbooks have been sold to businesses also.

Why should I get this instead of an Ultrabook that has a bigger screen.

If you're asking that, then why would you bother with an ultrabook in the first place? If you don't care about portability, then why not save $500 and get a non-ultrabook? Are you really so clueless that you don't see the advantage of a smaller form factor? You own an iPad, clearly enjoying the form factor for surfing, yet see no use in having that comfort when you bring just one computer with you while traveling?

...besides neither one is pocketable so why does it matter? You'd still have to carry a slate around.

iPads, MacBook Airs, and Ultrabooks aren't pocketable either, so why do they matter? Your logic is so full of holes that it's mind-boggling.

 

[Laurent06]

A lot of the poor experiences people had with the netbook was due to the single core Atom and slow HDDs (often 4200 RPM or really bad stuttering SSDs). Give Clover Trail a modern 64GB SSD and it'll be at least as fast as any ARM tablet at tablet tasks, and that's all it needs. The single core Medfield is already handling browsing and Javascript faster than the quad-core SIII. There's a million windows apps out there that don't need Core i3 speed.

A million apps that will work correctly and without having to use mouse/keyboard?

As far as speed goes, I'm afraid you don't know how Sunspider works: it isn't multi-threaded. Add to that Intel has been tuning its Android implementation for 2 years and that Linaro by just recompiling properly Android got a 30% speed up on Sunspider and suddenly Atom stands where it should: behin Cortex-A9. Also Sunspider is made of micro-benchmarks so that Medfield can turbo all along to 1.6GHz.

For ref, the scores (according to AnandTech) are 1279.4 for 1.6 GHz Medfield vs 1424.7 for S III which has A9 at 1.4 GHz. That means that in the current situation (without Linaro work) A9 and Medfield already have the same performance clock for clock for a single core.

There really isn't anywhere for ARM to go aside from the Apple faithful. Displays now take up such a large chunk of power consumption that even if ARM SoC makers had access to Intel's fab technology (and they don't), it still would be unlikely to make a big enough difference in battery life (or weight) to be a selling point.

Can you point to a study that shows that display eats "such a large chunk of power"? I have failed to find any recent study and I'm wondering.

Another AnandTech review of the iPad 2.4 seems to show that an iPad 3 is at about 4.9W for video playback while it goes to 7.6W on Sunspider and 10.3W on Infinity Blade 2. That seems to hint that even if the retina display indeed consumes a non negligible amount of power, the SoC itself consumes about the same amount, so I would conclude (but again, I'd like to see a serious study) that the SoC power consumption still is extremely important for a modern tablet.

 

[Mintmaster]

A million apps that will work correctly and without having to use mouse/keyboard?

Surface shows manufacturers how to make good keyboard covers, and a stylus is a very good substitute for a mouse. I actually think that this is why Apple shunned the stylus, because it lets you run any app designed for a mouse, and Windows has a huge advantage in that class of software. Apple wanted to start a fresh ecosystem. Aside from that, the stylus fits perfectly with Apple's "creative" image, iBooks strategy, targetting of highschool/college students, etc.

As far as speed goes, I'm afraid you don't know how Sunspider works: it isn't multi-threaded.

From Anandtech:
The SunSpider benchmark isn't explicitly multithreaded, although some of the tests within the benchmark will take advantage of more than one core.

Unfortunately, it's very difficult for me to find any comparison of single vs dual core with the same processor and OS. Here I can see that a dual core 1GHz A9 is faster than a single core 1.5GHz Scorpion, and you can only attribute a bit of that to architecture. Quad core A9's match dual core Krait (HTC One X vs One S, equal clock speeds), despite Krait having far superior IPC. Also, it's ridiculous for you to suggest that only Intel has been optimizing Javascript when we've seen big jumps with Android.

So yes, Atom is well ahead of A9 in Sunspider, and with CloverTrail it'll be more apparent.

Can you point to a study that shows that display eats "such a large chunk of power"? I have failed to find any recent study and I'm wondering.

I don't know of any study, but the iPad3 needs 4-8W, and even when gaming the total load only goes up a couple more watts over an idle home screen.

Intel is getting decent battery life in a phone, so it should be a small part of tablet power consumption.

Another AnandTech review of the iPad 2.4 seems to show that an iPad 3 is at about 4.9W for video playback while it goes to 7.6W on Sunspider and 10.3W on Infinity Blade 2. That seems to hint that even if the retina display indeed consumes a non negligible amount of power, the SoC itself consumes about the same amount

Yeah, but you're assuming that the device will be running heavy loads most of the time for the SoC to match the display consumption, which simply isn't a realistic usage scenario.

I doubt that ARM could get more than a 0.5W average advantage in battery life tests over Clover Trail, and I suspect it'll be substantially less. That's not enough to be a big selling point in a tablet.

 

[Laurent06]

From Anandtech:
The SunSpider benchmark isn't explicitly multithreaded, although some of the tests within the benchmark will take advantage of more than one core.

I invite you to run it on your PC and see what's going on. On all machines I tested, the benchmark itself only uses one core. I don't understand what AnandTech means, except perhaps that obviously having more than one core helps because OS tasks can be run on the other core; but in that case Medfield HT should be enough to take care of that.

Unfortunately, it's very difficult for me to find any comparison of single vs dual core with the same processor and OS. Here I can see that a dual core 1GHz A9 is faster than a single core 1.5GHz Scorpion, and you can only attribute a bit of that to architecture.

Snapdragon is much slower than A9.

Quad core A9's match dual core Krait (HTC One X vs One S, equal clock speeds), despite Krait having far superior IPC.

My understanding is that both One X and One S exist in both Krait and quad A9 variants so one has to be careful when reading results. I'm completely lost at deciphering which is using what Can you link some benchmark?

Also, it's ridiculous for you to suggest that only Intel has been optimizing Javascript when we've seen big jumps with Android.

Re-read my message please and what I said about Linaro. And I know pretty well what has been done by Intel, Google and ARM about all of that.

So yes, Atom is well ahead of A9 in Sunspider, and with CloverTrail it'll be more apparent.

You can keep on denying what I said, I have no issue with that. So let's talk again about that later this year, when Clovertrail will be out, OK?

I doubt that ARM could get more than a 0.5W average advantage in battery life tests over Clover Trail, and I suspect it'll be substantially less. That's not enough to be a big selling point in a tablet.

The issue with Medfield is not that it didn't get good battery life, in fact it's good enough. But they had to put a meager GPU to compensate the CPU power consumption. Any way in a tablet I agree with you that'd be totally different, so let's wait for tablet using Clover Trail and let's see what reviewers find for battery life.