ARMdevices.net

Dear Charbax,

I am big fan of ARM. My dream are silent ARM netbooks/notebooks,
however start of quad-core ARM will slow except nVidia

this is meaning: no way for 25% market share this year

My prediction:  3-7% this year,  17-30% in 2012

Charbax, You are right.

For me was absolute mystery extreme slow start of nVidia Tegra 2, until I understand:
Development of Windows CE *7?* suitable for dual-core ARM utterly failed and never was mentioned.
nVidia was without OS during 2010.

A number od semiconductor designers are planned quad-core CPUs around start 2012 (sampling end of 2011) – so this is no way for 25% market share in laptop area in 2011.

Last 5 year I was totally sick from Intel CPUs for laptops. Costly heating elements with NEVERENDING NOISE. Headache noise. When some folks are sitting at PC 10-12 hours/day, this Intel noise is something like torturing of Goyim.

This is my number 1 reason for ARM:  fanless heavenly SILENT !!  And no support of Apartheid II. – because Intel has 3 manufacturies in the so called “Israel”.

Sorry but just to get 25% of the laptop market would require ARM to completely replace all netbooks and they’re nowhere near ready to compete against the more powerful laptops.  Especially since it takes time for people to warm up to new products when they’re not sure they will be able to do everything they want it to do.

Though it is possible for them to co-exist.  Products like the CUPP computing PunkThis module are examples of how ARM and x86 could be combined for best of both worlds.  So while ARM is unlikely to make such strong market take over for laptops, they might reach such numbers in terms of hybrid systems. 

But Intel is going to aggressively improve their own power efficiency and performance range.  So that may not be a permanent arrangement once Intel and possibly AMD get into positions where they can start countering ARM advantages in by 2013-14.

However ARM is more likely to make strong progress in other markets, such as embedded systems, set top devices, etc.

Better business model is debatable, lack of legacy support and such quick planned obsolescence cycles may be fine for mobile devices like Smart Phones but doesn’t go over well with more expensive and less disposable products like laptops.  Really, the only types of people who really constantly get new laptops are either early adopters and/or gamers.

There is also the problem with product fragmentation with everyone doing their own thing with the technology, which is one of the reasons why MS is putting limitations and regulations on ARM systems designed for Windows 8.  While it is still a question of whether Windows 8 will provide legacy support and that could impede ARM adoption as well.  After all, there are still many x86 users that are still using Windows XP, especially on low end and/or old hardware, and won’t be easily tempted by an OS that forces them to start from scratch.

While most of the laptop market is still laptops!  Netbooks never ever really took over, they just had a higher market growth up till they reach market saturation for awhile.  Besides, most predict it’ll be desktops that will have their market share eaten into instead, with more and more people going for the mobile lifestyle.

Sorry but netbooks never actually reached 50% of the laptop market.  You’re using exaggerated market prediction numbers that never came true and are as erroneous as the premature claims that netbooks are dead.

The desktop wallpaper limitation only exists with Windows 7 Starter Edition and even that can be worked around. While Samba can be used on netbooks, and all you have to do is know how to configure it for sharing!  Even the original Eee PC with Xandros could run Samba!

Really, where are you getting your facts from?

The only thing you got right there is MS and Intel put limitations and the primary reason was to protect their higher profit products but also because such limitations helped ensure that netbooks would be cheap as possible to make with existing technology.  But such limitations were a luxury from when Intel had no competition in that market range. So both Intel and MS are changing policy and their next gen offerings will offer a lot more as they strive to fight off the new competition.

While no limits would not necessarily have allowed netbooks to take a much larger market share.  Not everyone likes using a mini-laptop for one thing, among other reasons if you really bother to think about it.

Thin clients have also been available for quite awhile and have never really taken off anywhere besides businesses for cost cutting.  Mostly because thin clients are pretty limited in performance over just a remote network connection.

While even if it did an explosion of users suddenly all using remote servers are going to use up all the bandwidth and cripple performance even more.

Intel is actually ahead of ARM in remote access anyway, with some of their latest improvements making possible for pretty much everything the remote client system is capable of.

Lower costs and power usage are also what Intel and AMD are going for, and good enough applies to them too to compete with ARM.

Like AMD Fusion still isn’t as good on run times as Intel ATOMs but they are viewed superior because of the APU and superior graphics.  Or in other words because they hold a performance advantage, while run time is good enough.  So the same would hold true if Intel significantly increase the ATOM performance while keeping cost and power usage within the good enough range compared to ARM.

Already Intel is pushing for some of the same advantages as ARM, such as instant on, always updated, etc. So ARM won’t be holding onto all of their advantages for long.

While already pointed out the ARM business model for mobile devices doesn’t necessarily translate well into the laptop market. 

Most of Apple’s profits came from the iPod before the success of their iPhone and then iPad models, but the iPod still represents a big chunk of their total earnings.  While their combined laptop and desktop shares have remained at only about 10% of the market.  And Apple is greedy and don’t share their profits with other companies. They’re also a closed and proprietary.  So not a good comparison to the rest of the ARM market.

You have to understand all business models have their strengths and weaknesses.  So you can’t automatically assume one will be superior in all cases.  Different market segments emphasize different traits that help products either succeed or fail in those market segments.  But a strength in one market segment can become a weakness in another.

Though I think you should listen to Sasha from Netbooknews.com on how he considers market analyst who try to predict more than a few months into the future.

Companies like Dell have been willing to lose those discounts and have been penalized for doing so.  Remember the rules use to be much stricter when Intel and MS first established those limitations.

So Intel and MS did handicap netbooks but the market is the market, they can only influence it and can’t prevent any product from dominating if that’s what the public really wants.

While AMD Fusion started with low end offerings precisely because they are interested in the netbook market and are also interested in getting into the mobile market.  AMD just didn’t have a viable product to compete with ATOM low power usage until they were able to develop their AMD Fusion line.  So don’t confuse them pretending not to be interested before with what they’re actually doing.  They even fired the guy who helped them get their Fusion line out in at least part because they felt he wasn’t pushing them towards getting into the mobility market fast enough.

Your reasoning for why lowering cost and power usage for Intel and AMD is also erroneous, in fact Intel is already on record stating they’re doing exactly that!!!  Their goal is to more than halve the existing TDP of their entire line up in the next two to three years.  While the bloat has been a fault of MS and not Intel, which at worse has only been compliant in as far as it was in their self interest.  But MS has been reversing this, first with Windows 7 that was less bloated than Vista and will again with Windows 8 that will be even more light than Windows 7.  So previous policies are no longer a valid argument.

The Intel ATOM in fact has been put on the same 2 year cycle as the rest of Intel’s processors, instead of the previous 5 year cycle they had it on.  The next update code named “Silvermount” will not only update the ATOM line to 22nm but also introduce a complete architectural reworking that will introduce Out of Order Processing, Intel’s 3D transistor technology, among other enhancements.  So while previously crippled, they’re rapidly pushing to make it relevant again.

You’re also wrong on previous thin clients having nothing to do with future ARM based thin clients. The same things that kept them from wide spread use before still apply.  The solution is flawed because it doesn’t provide all of the performance of running those operating systems on native hardware.  Only Intel is in a position right now to provide better remote performance. But you also ignored the limits of wireless bandwidth that is another flaw of relying on thin clients. Never mind that providers of thin clients usually charge for the service!

Fact is thin clients are already in use on ARM devices but they’re not being widely used outside of businesses, and the cloud has its limits.  So don’t expect arm to eat as far as you’re thinking into the rest of the PC market share.

I already clearly explained how Intel and MS imposed limitations, and that preventing loss of profit was the main concern but not the only concern, but not all companies complied to those limitation and no matter their influence they can’t control the consumers.  At best they can only purposely make a product less appealing.

While Intel is already reducing prices.  It’s happens to go along with Moore’s Law btw.  The 32nm Cedar Core ATOM’s go for about half ($43-$47) what the present Pine Trail dual cores ($86) go for.  Never mind netbook are pretty much the only major PC product that is sold near actual manufacturing costs.  So they were making less profit per chip since the beginning and is why they imposed limitations in the hopes that it would not eat into their higher end offering that still got priced with healthy profit margins.

But if they halve the cost again when they go 22nm then they become price competitive with ARM solutions.

Next, thin client have not been redefined!  It has always been a computer or computer program which depends heavily on some other computer (its server), whether you access your own or others (cloud) is irrelevant, to fulfill its traditional computational roles.

This encompasses both traditional thin client solutions and newer solutions like Chrome OS.  But systems that can leverage virtualization, to reduce the load on bandwidth (much like how buffering can smooth streaming video), can handle the interface better for more real time performance.

Neither VR or thin client though are true replacements for having that level of performance natively.  Citrix or other company solution all are limited to connection bandwidth.  While even if there was enough bandwidth to allow everyone to switch to thin clients, which there isn’t yet, the so called mobile broadband isn’t fast enough to handle even the video bandwidth requirement without serious data compression and less than HD resolution.

Services like OnLive require true broadband speeds to get anything approaching HD resolution.  While even 4G is too slow for HD.

And AMD did not spend tons of R&D money for over three years to develop the Fusion line and produce ultra low power solutions because they weren’t interested in the netbook market!  Everything from pricing to power efficiency puts them in direct competition with Intel ATOM.  But they won’t release anything capable of challenging Intel’s higher offerings until later.

AMD was simply not ready to get into direct competition when netbooks first came out.  So they feigned disinterest and for the first few years they settled for producing netbook alternatives that were between netbooks and CULVs performance with their AMD Athlon II Neo series.  All the while they were developing the AMD Fusion line, which they finally released in the beginning of this year.

The Bobcat CPU cores are very energy efficient but they’re less powerful than the previous Turion series.  They purposely tweaked it for greater power efficiency!  So it’ll take the upcoming Bulldozers with Llano APU for them to compete with anything more than Intel ATOMs.

While they have already tweaked the C-50 to produce their new Z-Series Z-01 to reduce TDP to ~5.9W for use in tablets and compete against Intel’s Oak Trail Z-Series solutions.

And no, AMD Fusion based systems are not priced significantly higher than Intel ATOM’s!

The Ontario single core 1.2GHz C-30 systems are going for about $280 (cheaper than the typical Intel single core solution), and dual core 1GHz C-50 is going for about $330 (equivalent to what Intel N550 systems are going for with premium models like the 1015PN with Nvidia ION going for even more, unless on sale since they’re presently pushing systems with the newer 1.66Ghz N570 and want to get rid of the older models).  While the Zacate often starts around $400, the Eee PC 1215B is only $436.99 on Amazon for the E-350 version with USB 3.0 port for example.

So sorry but you’re either using some really bad examples or you’re not really paying attention to the x86 market and are too focused on just the ARM market.

Charbax, please try paying attention to what I actually state.  I never said netbooks started out open, I clearly stated that both MS and Intel instigated limitations.  I was just more thorough and corrected you on the things you either missed or ignored like pointing out that some companies like Dell did violate those limitations and were penalized for it by temporary losing their discount rate.

Also for your information those limitations have been eased over time.  Netbooks can now officially be accepted in a size range from 7″ to 12.1″ in sizes.  Recommendation for SSDs now go up to 32GB or 250GB HDD (320GB is tolerated). Netbooks can be sold with 2GB of RAM as long as they are not sold with Windows 7 SE.  The new Cedar Trail N2800 is suppose to support up to 4GB of RAM now. 

Everything I stated about the changes for 2013 22nm Silvermount and halving their entire processor line’s TDP are straight out of what Intel themselves have been stating.  While enhancements like instant on are also being leveraged for Windows 8 with the new Intel systems that will be equipped with built in SSD.  Along with boosting HDD performance with installed SSD. 

You obvious missed the fact I pointed out netbooks are sold with far lower profit margin than the rest of the PC market, which includes most ARM devices!  So they’ll never be sold at the same price as Intel’s higher end offerings unless the cost of the system exceeds the cost of the higher end system.  Unless they either designed it to be a high cost ultra thin, and/or it uses more expensive parts like SSDs.

Cost and power usage for Intel also goes down as they reduce manufacture size.  Like I already pointed out even the 32nm Cedar Trail Intel is coming out with this year is basically halving the ATOM chip cost, along with getting the TDP as low as 3.5W, only half a watt higher than Oak Trail, for the 1.66GHz N2600 and all Cedar Trails will be dual core. 

The N2600 thermals are also low enough to even go fan-less and the higher end N2800 only takes up the TDP to 6.5W, which even adding the 2.1W from the NM10 south bridge still puts system power usage to about the present Intel N550/N570, which is 8.5W just for the ATOM and not counting the NM10.  So there is as much as a 2W improvement in power usage. 

Consequently, if they can halve the costs again with 22nm Silvermount then they’ll be in the range to be price competitive with ARM and start using sealed system designs to further save cost once fans are no longer needed.  It’ll just be a question which solution meets the needs of the consumers the best by then and whether Intel comes through with all their planned upgrades.

And again no, the thin client has not been redefined.  You’re confusing improvements in technology with the definition, which I repeat is a computer or computer program which depends heavily
on some other computer (its server), whether you access your own or
others (cloud) is irrelevant, to fulfill its traditional computational
roles. 

Exactly how this is done is irrelevant to the definition as the definition doesn’t cover how it works, only what it does!

And Intel didn’t exactly create the Netbook, the origins actually go all the way back to the networked computer concept of the mid-1990′s.  While subcompact laptops have been introduced as early as 1997 with Apple’s eMate 300.  But like tablets most such attempts failed to get much public interest.

Until the Eee PC finally kick started the netbook craze (name itself was created and became popular by the public) with just a Intel Celeron 353M Dothan chip. Intel just took advantage of that success to then impose limitations and introduce the ATOM that then took over the netbook market. 

Though the OLPC project, started about a year before, in 2006, the Eee PC came out, in 2007, was one of the major contributing factors that helped lead to the netbook’s final design considerations and market success.

Now while the Intel ATOM successfully cemented the start of the netbook market, they had no real competition and since it was low profits they put it under low priority with a 5 year update cycle.  Meaning they’ll add new tech innovations only after they had been thoroughly bug fixed and easier to manufacture.  The Pine Trail update to the ATOM pretty much only merged the CPU with the North Bridge for example.

But in this you are right, ARM has advanced tremendously in the last couple of years. 

While AMD has also entered the low end market and wishes also to enter the mobile market, they’re going 28nm next year and are basically racing Intel for 14nm by 2014… 

So Intel now has competition from both sides and they’re taking that threat seriously. Like I already pointed out that with Cedar Trail they will switch the ATOM to the same two year cycle as their higher end processors.  Cedar Trail is already getting features that were previously limited to just the Intel Core series with things like Intel’s WiDi for wireless N based streaming with compatible devices, standby times are being significantly increased from existing ATOMs.  But all that is pales to what they got planned for Silvermount in 2013 with not only going 22nm but adding more advance features like their new 3D Transistor technology, out of order processing, quad cores, etc.

Whether Intel or AMD will succeed in countering ARM or in the end decide to join them remains to be seen but don’t think for a second they won’t go down without a fight.  Intel didn’t get to where it is now by being push overs and AMD has survived some serious set backs in its history as well.

Btw, you may be getting Fiber soon but where I live I’m not getting it for probably another decade.  So that points out another example where you’re basing your opinion on something that is not a reality for all of us.

I get that you’re enthusiastic and a strong believer in ARM potential. I see that potential too, in fact I use both ARM and x86 devices, but I also know it’s not going to be as easy as you think it’ll be.

A) Charbax, I’ll repeat it for the last time, Thin clients have not been redefined!  A browser that can also function as a Thin Client doesn’t change what a Thin Client is!  All that is marketing BS you’re spouting that has nothing to do with the actual definition.  It’s like Intel promoting Ultrabooks, which is also mainly just marketing BS for what was already CULV range products.

Besides, Thin Clients that can sync with the remote system so you can access data offline are nothing new. 

B) While sure, Intel can play dirty, lots of big companies use such methods, but doesn’t change that they’re also improving their technology and it’s going to be a lot harder for ARM to go against than you’re predicting. Though playing dirty also means they got other advantages you shouldn’t ignore either.

C) While OLPC gets a lot of credit for contributing and applying pressure that helped bring about netbooks but you’re dreaming if you think netbooks wouldn’t exist at all without OLPC. 

At most it would just have taken longer and we probably would have ended calling them something else, as what OLPC helped do was pool ideas that were already out there and that helped start to focus those ideas for faster development.  Even the AMD Geode used in the first OLPC came out in 2002 for embedded systems four years before OLPC was even founded.

While it is true that the status quo was for faster and more expensive systems.  They were already reaching market saturation and were on the look out for ways to stimulate the market and develop new products to appeal to growing trends like the demand for easier to carry computers.

Like I already pointed out before this trend started all the way back to the mid 1990′s.  While OLPC was more of threat only to Intel’s Classmate PC project.  Since many of those kids you mentioned lived in countries that didn’t even buy Intel products because they couldn’t either afford them or couldn’t provide the power to use them.

It was actually the success of the Eee PC that convinced other companies to go all in to producing netbooks for the main market instead of just the poorer markets like OLPC was targeting and Intel was able to cash in on it before either VIA or AMD got a chance to.

So saying OLPC are the ones who made netbooks possible is like saying Apple invented the tablet when they did no such thing. They helped in a big way to make them mainstream faster but we were headed that way anyway!

Really, you shouldn’t believe everything OLPC says about the effect they had.  Like they claim they’re the reason why Windows XP is still around when the fact is Vista sucked and even companies that used desktops stuck to XP for years.

The only thing they really helped push that may not have happened otherwise was the extreme lowering of prices as companies are loath to reduce their profit margins.  But even that wouldn’t have happened if the companies hadn’t been convinced that would have helped them in the market to get more total sales and thus more gross profit.  So it’s the success of the Eee PC that really got Intel into it.  Otherwise it would have remained a niche product like all the previous mini PC products made before it…  Like the Toshiba Libretto 50 Ultra Mobile PC from 1997 for example, a full decade before the first Eee PC was released.

While back to the main point, you are ignoring that Intel has a strong established base.  ARM will need to be more than just on par to knock them down. Also technology like Intel 3D transistor technology is not easy to match.  You forget Intel probably has that well patented by now and other similar technology may not be as efficient.

You’re also ignoring that ARM has its own weaknesses, and again market advantage in the mobile market doesn’t mean they will continue to be advantages in the PC laptop market.

Really, ARM will wipe the floor with Intel on the Mobile front because that’s where they are the strongest and it’s their home turf.  But PC laptops are Intel’s home turf and where they are strongest.  It’s enough to say it’s impressive that ARM will challenge them but it’s fantasy to think they will dominate them on their own turf the moment they decide to enter it.  At best it will take a few years at the very least before they start seriously eating away at Intel’s dominance.

Sorry Charbax but all you’re doing is arguing is a revisionist and conspiratory view of history.

Like the fact mini-laptops have been around for over a decade, yet you want to make Intel into some conspiratory force that conspired with Asus to invent the Eee PC when the fact is companies have been introducing those types of systems since at least 1997. I even gave you a few examples. While Intel didn’t even introduce the ATOM until the Eee PC was already successful. 

You just don’t want to accept that all this was a natural evolution of laptops and make Intel out to be some evil force that ARM has to save us from is all exaggerated nonsense.  OLPC just accelerated things and the Eee PC’s success had a lot to do with timing and the slowing economy that made people more interested in getting low cost solutions.

But people want more than just low cost, which is why even $99 ARM netbooks never succeeded. 

The AMD Geode also wasn’t better than Intel, because it sacrificed even more performance and AMD didn’t stop developing the Geode as they were up to the 7th generation by 2007, they started since 1999!  The problem it always faced though wasn’t worry about disrupting the existing market, but the fact they couldn’t do it without sacrificing performance.  Even the early ARM offerings were barely useful for basic devices, let alone anything approaching a PC. 

Intel may have been a little worried for long term but they were a long way from that being a serious threat.

While Vista’s failure is one of the things that actually helped usher in the decline, not the end, of bloatware.  Software makers will always try to add on as much stuff as they can manage for their own financial interests but Vista showed MS they had to stop going in that direction and like pointed out it wasn’t OLPC or netbooks that made Vista a failure, it was Vista itself.

Another reality point is that products sold to the public are always a balance between what the companies want and what the public is willing to pay for.  The moment that dynamic changes then the companies have to adapt.  So the impact of the economic crisis had a lot to do with the sudden move to cheaper devices.

While the ARM market has actually been following the same pattern of performance and software bloat as the x86 market, with ever increasingly more power OS.  Just look at how Android has changed from when it first came out now to Honeycomb.  And now they newer alternate OS choices, including Windows itself!  So they are actually becoming more similar instead of being different the closer ARM gets to directly competing with x86 in the laptop market.

So Reality is ARM does have a good chance to dig deeply into the server market and the embedded device markets where they have the strongest argument to succeed but the laptop is a completely different animal and ARM actually has things going against it for that market that you are refusing to accept.

Most people don’t want to get new laptops every single year, most people want their laptops to be able to do everything they could possibly want it to do and not upgrading every year means they need extra performance to cover them and keep their system viable for the few years before they are ready to upgrade.  People also want as wide range of options for software they may want to run and only x86 will let them run everything and still run smooth without a performance hit due to emulation or running things remotely.  Even if Windows 8 can run legacy software it won’t do so as efficiently as it would on x86 hardware.

So no matter what x86 systems will continue to offer far greater performance than ARM, which will continue to only provide low end solutions that are only cheaper and more power efficient than x86 low end solutions.  While the cloud and emulations will not be sufficient to convince most people yet to give ARM a chance when they know the x86 systems can run the software without so many work arounds or have to worry about the performance hit of a emulator being too slow.

The market inertia is another factor that will continue to make most people buy x86 laptops because that’s what they’re use to as well.  It always takes years to get people to change gears and start using a different product from what they are use to, and lots of people still don’t trust the cloud completely.

While ARM is also waiting on Windows 8, to finally give them a powerful OS that can run more complex software and finally provide traditional desktop OS benefits for content creation and not just content consumption.  The other alternative OS’s that will start offering more than Android or iOS are also geared mostly to content consumption.  So while they wait on Windows 8, Intel and AMD are both working to improve their low end offerings to close the price and power efficiency gap and if they succeed then there will be far less to convince people to go with ARM instead of x86.

Part of the reason Windows XP survived so long is also because people are loath to give up using software they are used to using.  Like one of the biggest complaints against MS Office is they keep on changing the interface.  So if Windows 8 doesn’t fully support legacy software on ARM then that’s another hit against ARM for the laptop market, where the touch UI won’t make a difference, and HTML5 and Javascript apps will take time to develop.  While I already pointed out emulation and the cloud aren’t sure fire solutions to bridge the difference.

Intel as well would have to mess up and not deliver on their promised improvements to leave the way open for ARM to dominate even a quarter of the laptop market.  Let alone lose the market share you’re predicting.

Never mind ARM will eventually reach market saturation as well.  Everyone likes to dream a new market segment will revolutionize the whole market but all too often all it does is just create a wider and more competitive market.

Regardless, I’m just pointing out where you are being unreasonably optimistic and correcting you on what’s actually going on from your obviously biased point of view.  Not all your facts are wrong but you exaggerate many things and see more conspiracy than there actually was. 

I actually like the fact ARM will be challenging Intel, because competition is good and the end consumer benefits and agree they will be dominant in embedded market and in lots of new devices, in part because they will have no competition for those new device categories.  While there is no denying the server market is moving towards both ARM and Linux solutions. 

But no business model is perfect in every application and there are reasons why the laptop market is the way it is that can’t be ignored.  So there in is where we disagree and I have thoroughly explained why.  You can either accept it or ignore it, only time will show the final truth.

Don’t pretend you had access to any knowledge that wasn’t public Charbax…

Now you’re just being childish and trying to justify your stance by clinging to made up
facts from your revisionist view of history and the erroneous assumptions you have made from those views.

Actual history doesn’t support all your conclusions and you better start accepting that before you get labeled a fanboy loon.  Though some probably already put you under that category.

Many companies are just covering their bets right now, so don’t confuse that with a full scale migration.  The very same companies also supported other platforms before too.  Windows was even available for non-Intel systems before and is not a new trend.

You’re confusing your enthusiasm and hope for ARM with what’s actually happening and that’s causing you to exaggerate where they will be strong with markets where they still have far to go before they do more than pose a challenge to Intel on their home turf.

Really, instead of accepting the points that we agree on you are focusing on forcing that your entire dreamed up future will be the actual future we will see.  You can’t seem to accept even the possibility that you could possibly be off the mark and that seems to be all ego on your part.  So try thinking on that before further wasting our time with this continued diatribe on your part.