They are paying a lot of money ($32 billion) for gross income of around $1.2billion, and profits of about 1/2 of that. Based on that, it will take 30-40 years for the investment to pay SoftBank off, based on ARMs revenue alone.
Something doesn't add up. On one hand, they're saying that they don't plan to change their business model, or even how their R&D is done in the next years, but on the other hand - they will need to come up with something really amazing if they want to recover this investment.
You can buy a lot of other things for $32 billion that would be a far safer bet.
ARM and Intel do not compete directly. At least not financially.
Intel's competitors (for example, in the mobile or embedded businesses) are the likes of Qualcomm, Broadcom, ST and NXP.
Intel sells discrete CPUs, while everyone else licenses a CPU core from ARM, adds in their own stuff, and then sells that. As long as someone needs to add some extra stuff in there, ARM is still the industry's preference.
Things may get interesting with the recent acquisition of Altera (by Intel), but there is still a big gap between the more "generic" processors that Intel makes and the application specific chips that ARM licensees are using.
Chinese manufacturer's make products, which are sold to customers in the US (in this example), in US Dollars. But these aren't much good to the manufacturer (i.e. he doesn't really want to pay his employee's in USD, or at least the Chinese government wouldn't approve of that), so he sells the USD to the People's Bank of China, who use these dollars to buy Treasury Bonds. In return, the People's Bank gives the manufacturer the amount of Yuan that it thinks is appropriate.
Nokia of course will continue internal development of Android as a software platform for their phones and have an escape clause for when M$ deal goes tits up, they aren't that silly.
Sorry, but Nokia is not (and never was) developing anything Android related. Nokia's software strategy for smartphones were Series 60 Symbian and Meego.
If you have C/C++ code, porting it to ARM should be a huge deal. Yes there will be some differences, yes, there will be bugs - but in terms of effort its manageable. And more importantly, every single vendor has to do this effort, so Microsoft doesn't have to do anything.
Because Microsoft are saying this now, with no product that anyone can "buy" right now (or even soon), this probably means the audience for this news is *Developers* (the single intelligent word that Mr Balmer has uttered in the last 20 years, so good, he had to say it multiple times for it be considered a quote). They are now selling the ARM architecture to developers. If the developers buy this story, the applications will follow.
And of course, some developers will be more prepared than others. Don't expect an ARM version of Photoshop anytime soon, but an ARM version of Firefox is something that could be cranked out very easily.
Wikipedia tells me that.NET (and therefore managed code) is nearly 10 years old (13 February 2002)
I'm pretty sure that someone in Redmond was thinking about supporting multiple platforms when they started architecting their software compiler strategy back then. It just took their management structure 5 years to wake up to the idea.
Now people have to go in and remove all of that crud which is blocking porting their SW to a different architecture...
DLL Hell was yesterday, tomorrow is P/Invoke hell.
This sounds like a glitch in the search algorithm than anything else. Publishing trends is interesting, and can allow us to learn more about what we (as a species) do with the internet. This information is clearly abused by a few (who then go out and write fake page which use the popular keywords to attract attention to their page), but this is an abuse of the Trends information that google provides, not something inherently evil.
Google (or any search engine) could just tweak their results to reduce the importance of sites which are written *after* a topic became trendy. At least to give the existing articles a head start. Or I can imagine a million other ways in which they could tweak the algorithm.
But I don't think what the article is implying (that google should stop publishing Trends) should be taken seriously.
From 15 minutes of research, its pretty clear to see that Brighthouse Labs, Molinker and FidesReef were (/are) definitely polluting the App Store.
In some cases, its very clear why Apple have introduced the "In App Purchasing" - most of these should probably disappear when all the duplicated Apps get merged back to a common one.
But from Apples point of view, they're still trying to prove that their platform has the most Apps. They've already put the "technology" into place - so now all they need to do is rush in and save the day, by forcing the developers to merge their Apps.
But seriously, with power comes responsibility. They're finding out that its not always easy to walk that tightrope.
I would turn the question around a different way: - what does Microsoft get out of helping Mono ?
(remember, software engineers probably cost MS 200-300K per year, in terms of cost, not salary).
While it looks like they are partially serious about it, I have a nasty feeling (and again, only my opinion), that we're in the "embrace" period of Microsofts three point plan. Darn, I can't remember what the next two parts are... i had it just a second ago, but i think the next parts are definitely not good.
Finally, on directions of Mono: I'm sure that Miguel's intentions are good but (i just had to dig up some Mark Twain here): Half the results of a good intention are evil; half the results of an evil intention are good.
That works for both Mono and Microsoft. Only time will tell.
But I guess at the end of the day, the real benefit of Open Source is *choice*.... and we still have it.
We can decided to use Mono, or we can choose not to.
Some people have already. And I truly believe that the work that Miguel and Novell are doing will entice more people to use it as a platform to program.
But we have to be a little bit cautious when dealing with the Beast (MS). Their track record isn't all that pretty !;-)
The Vt of the standard cells is not the same as the process option. You can get a "Low Power" process with low and high Vt cells, and likewise a "High Performance" process with lowand high Vt cells.
Less transistors switching is only part of the story.
Maybe a more signficant factor in determining the power consumption of a CPU is the technology process choice.
Intel typically tune their process for performance, at the expense of leakage. This lets them squeeze out a couple of GHz in terms of clock speed, but it means that the power consumed when the chip is doing nothing at all (i.e. idling) is much larger. The CPUs that are put into cell phones (from companies like ST, TI, Broadcom, etc, etc) are normally fabbed with a "low power" or LP option. This reduces the maximum speed that you can get out of the processor, but reduces the leakage problem significantly. If the cell phone is only using the processor 1% of the time (think of how long it spends powered on in your pocket), then there is no point in having the best 3D games on your phone, if the stand-by time is 15 minutes.
Switching between these standard (or GP) processes and LP processes is not quiet straight forward, as you need to design all your mixed-signal / analog blocks (think PLLs, bandgaps, regulators, etc) for both nodes. While I'm sure Intel could probably afford to do this, they would then have to turn around and support this process in their fabs, which would eat up their resources for their processor market.
If you compare the numbers: Intel can sell their processors for hundreds of dollars. Phone manufacturers buy processors from the other Semicos at about 10-15 dollars each. Guess where the better margin is...
While the author correctly identifies a huge potential market for smartphones in the coming years, maybe his assumptions about Symbian are a little naive.
These smartphones are becoming popular because they are becoming more and more like a standard PC every day. The only exception being the user interface (if anyone has an idea how to fix this, give me a call ! I promise to share in the huge profits ! ).
This is facilitated by the increasing processor power that these phones have available to them. Symbian was designed for small memory, low performance processors which incredibly strict power consumption requirements and limited connectivity running in a highly controlled environment (i.e. software environment).
The cost of developing drivers for Symbian (with all its quirks) is enormous. At the moment, the semiconductor companies are getting hit with the cost of this development. This will not last forever, they will always strive for the cheapest possible solution - and this helps explain Linux large penetration in this market.
The company that holds the best cards in this field is Apple. They have waited until mobile devices have become powerful enough to run (only slightly modified) standard PC kernels (XNU). This is going to save them a fortune in the years to come. Microsoft has missed this boat - they are trying to split their OS into as many different branches/versions/flavours as possible, while neglecting the requirement to try and maintain a common "brand" across all devices.
Are there issues with the current GPL that need to be fixed ? Or even some parts that need to be clarified ?
If there are actual issues with the license, then a rewrite is a good thing - all I'm concerned about is that people don't waste time developing a new license when one isn't needed. In the end, its adoption will be decided by the various projects - on a case by case basis, so just because there is a version 3, doesn't guarantee adoption, unless it brings benefits.... And hopefully it doesn't spawn pro- and anti- GPLv3 wars in every GPLv2 project ! While licensing is important, it shouldn't create huge overheads that distract developers from doing what they do best.
Why not use the bluetooth connection on your appropriately spec'ed cellular phone to download tracks from iTunes to your iPod.
This gives people the CHOICE to use whatever phone they want, with whatever MP3 they want with whatever Online service they want. (Okay, if you want iTunes, then the iPod is not optional).
Its a win-win for Apple - suddenly they do not need reseller deals with every Mobile operator in a country to sell their wares.
My television does not need to be able to wash my clothes and clean my carpet.
Of course there are some issues to be worked out, one of the biggest is that the main carriers want to control your life... I don't buy books/music/dvd's off my broadband provider, what makes the network operators (i.e. Verizon/Cingular/whoever) think that I should buy these from them. If they got back to doing what we're paying them to do - provide a cheap cost-effective data service - then they might actually be able to succeed in that goal.
You have 1,000s of customers with root on their own boxes. The FBI now wants the log files of customer 41231. Its going to take a tech at least 15 minutes (best possible case) to fetch the log files from this box - assuming that the log files that the FBI want are in their usual place. This is giving the tech enough time to log into the box, find them, and FTP them somewhere else.
Just imagine how long this could take if the customer has gone to some effort to make the log files difficult to find.
Looks like the ISPs didn't complain loud enough when these laws were introduced and they're now paying for it.
Could be wrong here, but I thought that the main reason for implementing a CMT chip with "hardware threads" was to make the context switch less painful.
On single processor systems, when it wants to switch between two threads, it usually executes a context switch - it needs to dump one set of registers to memory, load the other set from memory and change the instruction pointer.
That usually adds up to two seperate memory accesses to different parts of memory. What's more, is that it is not always possible to accurately predict (by the processor) which two sets of memory addresses will be involved - it all depends on what new thread has been choosen by the scheduler.
This wouldn't have been a huge problem on Intels architecture - given the relatively small number of registers it gives to applications.
In Sparc architectures - you have a lot more registers (and wider 64 bit registers) to worry about.
What Sun would like to do is remove this overhead by implementing a set of registers for each of processing unit. This makes them independant in their own right.
At the end of the day, the bottle neck in most systems is going to be the RAM-CPU bus, if this can reduce the number of hits that bus takes, then overall system performance will improve - by what margin is usually up to the system architects (i.e. why pick 8 cores instead of 12, why pick 3Mbit of cache instead of 2MBits, etc, etc)
It was a tough choice, but I doubt Apple moved to Intel for cheaper chips, or better processors. Intel has always developed chips that aren't x86 or IA64 for "research" purposes.
I'd imagine that Apple are probably after Intels vast fabrication resources. They probably see that IBMs fabs will probably be under pressure to crank out chips for the XBox and Playstation.
For the volumes of chips that those two platforms will need, its hard for IBM to justify Apple taking up their valuable fab space.
Slashdot Mirror
Well, i assume the %pCOW in printk.c is a joke?
But no one on slashdot reads the articles - so I guess itâ(TM)s pretty well hidden.
You must be new here ... Slashdot readers never bother to visit the actual link itself.
The fact that you're the first person to notice this and there are already 32 posts kind of proves this.
Can we rate this story flamebait ?
It looks like Slashdot (and Techcrunch?) is just trying to bring up the whole Dash debacle again to generate traffic.
Resist people ... resist!
Supposed to be a 1/5 th model of the Shuttle:
https://www.youtube.com/watch?...
The SRBs were 150ft tall, so 1/5 would make it only 30 feet. But there are two of them!!
This makes no sense for SoftBank.
They are paying a lot of money ($32 billion) for gross income of around $1.2billion, and profits of about 1/2 of that. Based on that, it will take 30-40 years for the investment to pay SoftBank off, based on ARMs revenue alone.
Something doesn't add up. On one hand, they're saying that they don't plan to change their business model, or even how their R&D is done in the next years, but on the other hand - they will need to come up with something really amazing if they want to recover this investment.
You can buy a lot of other things for $32 billion that would be a far safer bet.
ARM and Intel do not compete directly. At least not financially.
Intel's competitors (for example, in the mobile or embedded businesses) are the likes of Qualcomm, Broadcom, ST and NXP.
Intel sells discrete CPUs, while everyone else licenses a CPU core from ARM, adds in their own stuff, and then sells that. As long as someone needs to add some extra stuff in there, ARM is still the industry's preference.
Things may get interesting with the recent acquisition of Altera (by Intel), but there is still a big gap between the more "generic" processors that Intel makes and the application specific chips that ARM licensees are using.
"China" is a rather vague definition.
Chinese manufacturer's make products, which are sold to customers in the US (in this example), in US Dollars. But these aren't much good to the manufacturer (i.e. he doesn't really want to pay his employee's in USD, or at least the Chinese government wouldn't approve of that), so he sells the USD to the People's Bank of China, who use these dollars to buy Treasury Bonds. In return, the People's Bank gives the manufacturer the amount of Yuan that it thinks is appropriate.
By that definition, does a Pension also qualify as a Pyramid scheme ?
Nokia of course will continue internal development of Android as a software platform for their phones and have an escape clause for when M$ deal goes tits up, they aren't that silly.
Sorry, but Nokia is not (and never was) developing anything Android related. Nokia's software strategy for smartphones were Series 60 Symbian and Meego.
Well, this depends on their target audience.
If you have C/C++ code, porting it to ARM should be a huge deal. Yes there will be some differences, yes, there will be bugs - but in terms of effort its manageable. And more importantly, every single vendor has to do this effort, so Microsoft doesn't have to do anything.
Because Microsoft are saying this now, with no product that anyone can "buy" right now (or even soon), this probably means the audience for this news is *Developers* (the single intelligent word that Mr Balmer has uttered in the last 20 years, so good, he had to say it multiple times for it be considered a quote). They are now selling the ARM architecture to developers. If the developers buy this story, the applications will follow.
And of course, some developers will be more prepared than others. Don't expect an ARM version of Photoshop anytime soon, but an ARM version of Firefox is something that could be cranked out very easily.
Wikipedia tells me that .NET (and therefore managed code) is nearly 10 years old (13 February 2002)
I'm pretty sure that someone in Redmond was thinking about supporting multiple platforms when they started architecting their software compiler strategy back then. It just took their management structure 5 years to wake up to the idea.
Now people have to go in and remove all of that crud which is blocking porting their SW to a different architecture ...
DLL Hell was yesterday, tomorrow is P/Invoke hell.
Aha,
so more like a "we put our own money in there, so it has to be a good thing".
reading the article properly helped a little bit ! :)
Did evil just become even more evil while I was sleeping over New Years ?
Goldman Sachs aren't exactly known for their "good values".
Why the hell does an investment bank, who normally act as a "service provider" want to take a direct stake in a Social networking company ?
... great now this post will make this article appear as "fake" when you want to really find articles which are relevant to Britney and Lindsay ...
I'm sorry, but lets take a step back here ...
This sounds like a glitch in the search algorithm than anything else. Publishing trends is interesting, and can allow us to learn more about what we (as a species) do with the internet. This information is clearly abused by a few (who then go out and write fake page which use the popular keywords to attract attention to their page), but this is an abuse of the Trends information that google provides, not something inherently evil.
Google (or any search engine) could just tweak their results to reduce the importance of sites which are written *after* a topic became trendy. At least to give the existing articles a head start. Or I can imagine a million other ways in which they could tweak the algorithm.
But I don't think what the article is implying (that google should stop publishing Trends) should be taken seriously.
Or even better,
Check the stats on their own database (i.e. the App Store).
I found this just now: http://www.iappphone.com/stats/
The top 5 "Submitters"
Brighthouse Labs - 1855
Iceberg Reader - 1369
Molinker Inc. - 1011
FidesReef - 825
iLike inc - 588
From 15 minutes of research, its pretty clear to see that Brighthouse Labs, Molinker and FidesReef were (/are) definitely polluting the App Store.
In some cases, its very clear why Apple have introduced the "In App Purchasing" - most of these should probably disappear when all the duplicated Apps get merged back to a common one.
But from Apples point of view, they're still trying to prove that their platform has the most Apps. They've already put the "technology" into place - so now all they need to do is rush in and save the day, by forcing the developers to merge their Apps.
But seriously, with power comes responsibility. They're finding out that its not always easy to walk that tightrope.
Sorry, it was too tempting !
I would turn the question around a different way:
- what does Microsoft get out of helping Mono ?
(remember, software engineers probably cost MS 200-300K per year, in terms of cost, not salary).
While it looks like they are partially serious about it, I have a nasty feeling (and again, only my opinion), that we're in the "embrace" period of Microsofts three point plan. Darn, I can't remember what the next two parts are ... i had it just a second ago, but i think the next parts are definitely not good.
Finally, on directions of Mono: I'm sure that Miguel's intentions are good but (i just had to dig up some Mark Twain here):
Half the results of a good intention are evil; half the results of an evil intention are good.
That works for both Mono and Microsoft. Only time will tell.
But I guess at the end of the day, the real benefit of Open Source is *choice*. ... and we still have it.
We can decided to use Mono, or we can choose not to.
Some people have already. And I truly believe that the work that Miguel and Novell are doing will entice more people to use it as a platform to program.
But we have to be a little bit cautious when dealing with the Beast (MS). Their track record isn't all that pretty ! ;-)
The Vt of the standard cells is not the same as the process option. You can get a "Low Power" process with low and high Vt cells, and likewise a "High Performance" process with lowand high Vt cells.
See the bottom graph on: http://www.umc.com/english/process/a.asp
The "Core Devices" contain the standard cells. UMC call "SP" the normal leakage process, and "LL" the low-leakage process. TSMC have similar.
Less transistors switching is only part of the story.
...
Maybe a more signficant factor in determining the power consumption of a CPU is the technology process choice.
Intel typically tune their process for performance, at the expense of leakage. This lets them squeeze out a couple of GHz in terms of clock speed, but it means that the power consumed when the chip is doing nothing at all (i.e. idling) is much larger. The CPUs that are put into cell phones (from companies like ST, TI, Broadcom, etc, etc) are normally fabbed with a "low power" or LP option. This reduces the maximum speed that you can get out of the processor, but reduces the leakage problem significantly. If the cell phone is only using the processor 1% of the time (think of how long it spends powered on in your pocket), then there is no point in having the best 3D games on your phone, if the stand-by time is 15 minutes.
Switching between these standard (or GP) processes and LP processes is not quiet straight forward, as you need to design all your mixed-signal / analog blocks (think PLLs, bandgaps, regulators, etc) for both nodes. While I'm sure Intel could probably afford to do this, they would then have to turn around and support this process in their fabs, which would eat up their resources for their processor market.
If you compare the numbers: Intel can sell their processors for hundreds of dollars. Phone manufacturers buy processors from the other Semicos at about 10-15 dollars each. Guess where the better margin is
While the author correctly identifies a huge potential market for smartphones in the coming years, maybe his assumptions about Symbian are a little naive.
These smartphones are becoming popular because they are becoming more and more like a standard PC every day. The only exception being the user interface (if anyone has an idea how to fix this, give me a call ! I promise to share in the huge profits ! ).
This is facilitated by the increasing processor power that these phones have available to them. Symbian was designed for small memory, low performance processors which incredibly strict power consumption requirements and limited connectivity running in a highly controlled environment (i.e. software environment).
The cost of developing drivers for Symbian (with all its quirks) is enormous. At the moment, the semiconductor companies are getting hit with the cost of this development. This will not last forever, they will always strive for the cheapest possible solution - and this helps explain Linux large penetration in this market.
The company that holds the best cards in this field is Apple. They have waited until mobile devices have become powerful enough to run (only slightly modified) standard PC kernels (XNU). This is going to save them a fortune in the years to come. Microsoft has missed this boat - they are trying to split their OS into as many different branches/versions/flavours as possible, while neglecting the requirement to try and maintain a common "brand" across all devices.
Are there issues with the current GPL that need to be fixed ? Or even some parts that need to be clarified ?
... And hopefully it doesn't spawn pro- and anti- GPLv3 wars in every GPLv2 project ! While licensing is important, it shouldn't create huge overheads that distract developers from doing what they do best.
If there are actual issues with the license, then a rewrite is a good thing - all I'm concerned about is that people don't waste time developing a new license when one isn't needed. In the end, its adoption will be decided by the various projects - on a case by case basis, so just because there is a version 3, doesn't guarantee adoption, unless it brings benefits.
Another reason to hold on ROKR...
... I don't buy books/music/dvd's off my broadband provider, what makes the network operators (i.e. Verizon/Cingular/whoever) think that I should buy these from them. If they got back to doing what we're paying them to do - provide a cheap cost-effective data service - then they might actually be able to succeed in that goal.
Why not use the bluetooth connection on your appropriately spec'ed cellular phone to download tracks from iTunes to your iPod.
This gives people the CHOICE to use whatever phone they want, with whatever MP3 they want with whatever Online service they want. (Okay, if you want iTunes, then the iPod is not optional).
Its a win-win for Apple - suddenly they do not need reseller deals with every Mobile operator in a country to sell their wares.
My television does not need to be able to wash my clothes and clean my carpet.
Of course there are some issues to be worked out, one of the biggest is that the main carriers want to control your life
Just think about it ....
You have 1,000s of customers with root on their own boxes. The FBI now wants the log files of customer 41231. Its going to take a tech at least 15 minutes (best possible case) to fetch the log files from this box - assuming that the log files that the FBI want are in their usual place. This is giving the tech enough time to log into the box, find them, and FTP them somewhere else.
Just imagine how long this could take if the customer has gone to some effort to make the log files difficult to find.
Looks like the ISPs didn't complain loud enough when these laws were introduced and they're now paying for it.
Or more so, the customers are paying for it.
Could be wrong here, but I thought that the main reason for implementing a CMT chip with "hardware threads" was to make the context switch less painful.
On single processor systems, when it wants to switch between two threads, it usually executes a context switch - it needs to dump one set of registers to memory, load the other set from memory and change the instruction pointer.
That usually adds up to two seperate memory accesses to different parts of memory. What's more, is that it is not always possible to accurately predict (by the processor) which two sets of memory addresses will be involved - it all depends on what new thread has been choosen by the scheduler.
This wouldn't have been a huge problem on Intels architecture - given the relatively small number of registers it gives to applications.
In Sparc architectures - you have a lot more registers (and wider 64 bit registers) to worry about.
What Sun would like to do is remove this overhead by implementing a set of registers for each of processing unit. This makes them independant in their own right.
At the end of the day, the bottle neck in most systems is going to be the RAM-CPU bus, if this can reduce the number of hits that bus takes, then overall system performance will improve - by what margin is usually up to the system architects (i.e. why pick 8 cores instead of 12, why pick 3Mbit of cache instead of 2MBits, etc, etc)
It was a tough choice, but I doubt Apple moved to Intel for cheaper chips, or better processors. Intel has always developed chips that aren't x86 or IA64 for "research" purposes.
I'd imagine that Apple are probably after Intels vast fabrication resources. They probably see that IBMs fabs will probably be under pressure to crank out chips for the XBox and Playstation.
For the volumes of chips that those two platforms will need, its hard for IBM to justify Apple taking up their valuable fab space.