I'm not quite sure what the guy that wrote this article is smoking, but there is pretty much no way LinkedIn is going to push MSFT above the $1 trillion mark. Sure the only thing he claims at that MSFT will make $1 trillion before Apple or Alphabet, but at current growth rates none of those companies are going to be in the ball park of $1 trillion valuation for 5-10 years at least, probably longer. It seems unlikely that LinkedIn is the killer acquisition that is going to drive their growth for the next decade (as opposed to all the license revenue they make from existing established business.)
MSFT's current market value is $491.71 billion. Hypothetically, lets say that being acquired by MSFT does not change the value of LinkedIn at all, if that casepost-merger MSFT would be worth... $491.71 billion, they paid $26 billion in stock + cash for LinkedIn reducing the value of MSFT by $26 billion, then the value of LinkedIn ($26 billion) gets added to their market cap, cancelling out the two effects. Now lets say they double the value of LinkedIn by giving it access to their network of enterprise customers, they MSFT's value becomes $491.71 + $26 = $517.71 billion... a 5% increase in the value of MSFT.
I'm sorry, there is just no plausible scenario where LinkedIn suddenly becomes worth more than the entire rest of MSFT is worth. There is no way LinkedIn is going to double the amount of license revenue they generate. That's the problem that enormous businesses like MSFT have... for almost any other business an extra $500 million of quarterly revenue would be an enormous new windfall that would double their stock's value. For MSFT... it would be an extra 2% growth of quarter revenue, to which Wall Street would yawn and say "Here's your gold star MSFT, stock is up 2% after quarterly earnings." Its hard to create big new percentage growth when all the new technologies are not worth anywhere close to your existing massive business. MSFT's partner Intel has also been facing this problem for over a decade. Just like any other company, Apple and Google are not immune to this and are starting to hit the wall that the burden of massive valuation begets. Any company in this position more or less is forced in to paying dividends and becoming a boring "blue chip" stock. Nothing wrong with it really, take the example of 3M, a diversified technology company that has been a stable steady company. They have increased dividend rates every year for over a century now, and their stock holds strong value long term. I think the tech industry as a whole needs to wake up to the fact that we are rapidly turning in to that type of old but strong and steady business.
I'm pretty sure most of the problem is Adobe somehow managed to make the newest version of Flash even slower and more resource consuming than Flash from a year ago. But you absolutely have to make sure you have the newest version installed at all times, since Flash is also a security mess, despite the fact that Adobe has also made keeping Flash up to date a huge PITA (to the point that MSFT and Google handle the updates for them on their respective browsers.) The performance and security of Flash has been a continuing problem for decades now.
Why they bother to continue to develop and release new features for it puzzles me, plugin based content is very clearly a dying platform at this point and its partially Adobe's fault (we wouldn't want to get rid of Flash if it didn't suck so hard.)
I'm sick of these articles that sound like they are mansplaining the basics of tomahawk fusion that we have known since the 1970s and then claims its a new thing. Moreover, they supposedly have a working commercial reactor when we know that a commercial reactor would need to be ITER sized for positive energy generation. Can we keep this crap off slashdot?
The USB power delivery spec standardizes how to increase Vbus voltage and max current. Power profile 5 increases the voltage to 20V with 5A current for 100W of power. It can be implemented on either type A, micro USB, or type C. It used to be that the USB spec only standardized up to 7.5W power draw, which became a limiter on charging time. Now that we have the new power delivery spec extension, there are zero good reasons to implement proprietary charging standards to move beyond 7.5W.
Despite this, Qualcomm is still heavily marketing its proprietary quickcharge 3.0 system to smartphone manufacturers, purely for the incremental profits on licensed wall chargers. I'm glad to see Google throwing their weight around a little in an effort to shut down what is purely a money grab.
It looks like Intel is still playing catch up in the modem space. Interestingly, it looks like for the 2016 iPhone, Apple is using either the Intel XMM 7360 or the Qualcomm X12. Both of these modems were released in 2015. Qualcomm hasn't shipped a new generation since then, but Intel did release the XMM 7480 in February. It would be interesting to see how much progress Intel has made in a year.
Either way, the fact that Intel's modem exists is good for everyone... except Qualcomm. Without it, Qualcomm would be the only LTE modem supplier. There is no doubt Apple is aware of the Intel modem's shortcomings. My guess is Apple is willing to turn a blind eye to that for the "1st gen" product and use the new revenue as a carrot to get Intel to direct its engineering efforts to the features that Apple wants, probably stuff like CDMA for example.
Unless Apple is doing some sort of quasi RAID like read/write access against all available memory chips
"Quasi RAID" is actually how all SSDs work. The controller spreads the writes across multiple flash chips. This is why NVMe is so much faster, the OS can give the SSD controller thousands of outstanding IO requests instead of the max of 32 with SATA. The iPhone uses a single eMMC flash chip which integrates the controller and the NAND on one die. The eMMC chip will do the same thing, only across flash cells instead of entire chips.
In the end, all this comes down to is the fancy 128/256GB eMMC flash comes with a nicer onboard controller than the 32GB one, hence higher bandwidth. That said, with such a huge difference there is no doubt Apple ordered the cheapest 32GB flash they could find. You probably can find the same eMMC chips in a $50 cheap Chinese Android phone. For a $500+ phone they should be paying the extra $2 for higher bandwidth 32GB eMMC chips.
Remember this is Russia. Most of the USSR state owned industry was sold to former Communist party members and other politically well connected people for pennies on the dollar, including the Soyuz factories. It takes 2 years to build a Soyuz because they use the exact same industrial process to build them that they used 40 years ago under Communist rule. It's Govornment subsidized and NASA pays through the teeth for it anyway, so zero incentive to improve inventory management or process efficiency.
This Gartner report seems to tell a very different story than Intel's announcement that Q3 revenue would $700 million better than expected. To quote Intel's press release:
The increase in revenue is primarily driven by replenishment of PC supply chain inventory. The company is also seeing some signs of improving PC demand.
Intel can get in big legal trouble with the SEC for lying on these type of financial announcements, so I tend to give a little more cred to regulated stuff like this than just some random analyst.
Now before anyone freaks out, keep in mind that there is nothing in the Audio over USB-C spec that requires the device that implements it to not have a headphone jack. Its totally legit for a phone to support this spec and have a headphone jack. In fact, I suspect that most vendors will probably go this route.
Being able to plug your phone in to a single USB-C connector on your car or stereo and have it charge the battery and play music using 1 wire is a nice feature.
Really should read "UK's Top Police Warn That Making Aim-Bots/Game Cheats May Turn Kids into Cyber Criminals"
I'm not an expert in sociology, but it seems plausible that unethical behavior in online video games can be a gateway to unethical online behavior in general. From a technical standpoint I know that the skills developed by hacking games are similar to the skills needed to hack financial software.
I don't think this made the Slashdot front page but Intel bought VIA's CDMA modem design and license about a year ago. Intel's modems currently only support GSM & LTE, whereas VIA never updated their CDMA modems to be LTE capable. It likely will take a couple years for Intel to integrate VIA's CDMA implementation with their LTE design, but once its done, Intel's modems will be just as capable as Qualcomm's.
CMDA isn't only important for the US Verizon/Sprint market, the much more important reason to implement CDMA is China Telecom. Either way, the iPhone 7S will likely mark the return of all iPhones being universally supported by all carriers... regardless of whether there is an Intel or Qualcomm modem inside it.
No word from Apple about whats going to happen with the Apple Watch Edition. Seriously if you are going to sell a $17,000 luxury item that will be obsolete 16 months after it is released you better have a good upgrade plan in place.
Its a little disingenuous to say that Watson "created" the trailer. The only thing Watson did was run a pattern recognition algorithm to figure out which clips in the movie were tense, happy, scary, etc. Then a human editor sorted through all of the clips, picked the good ones and put them in sequence to create a trailer that actually had narrative instead of just being a hodge podge of disjoint clips.
Pattern recognition is getting better which is the first step to creating an AI... but Watson, and AI in general is still very far off from creating a computer program that is capable of original thought.
Java is now toxic thanks to its owner. For the sake of the entire tech industry, we all should consider it a legacy technology that should be removed from everything as quickly as possible. Unfortunately that will take years... maybe even decades, but we must start the deprecation process as quickly as possible. Besides, in the 20 years since it was created we have better cross platform languages now anyway.
Thankfully a lot of us have input in to technical decisions here. We all need to take a stand and kill Java.
I keep a Windows system around for minor software that needs it
AKA "games".
Other than games, the very important thing that keeps Windows on my personal system is TurboTax. Like pretty much any other US Taxpayer that has a tax situation too complex for form 1040-EZ and doesn't want to pay ~$150 for H&R Block or ~$300 for a certified CPA. I hired a CPA once and $50 per year TurboTax did a better job!
Before anyone says Wine, its a non-starter. TurboTax uses a bunch of.Net features that don't work 100% right on Wine like WPF. Unfortunately the Mac version is absolute garbage, so that route isn't viable either. It really sucks, but the easiest way to be a lawful US citizen is to have a Windows system.
Trying to argue that Chrome isn't trans-formative when compared to Java SE is so ridiculous its almost funny. You can only run litigious business like Oracle based on making everyone else pay a protection racket for so long.
The fact that Intel is offering to manufacture ARM cores for their custom foundry customers is not new. In fact, there are some Altera FPGAs with embedded ARM cores being manufactured by Intel already. The important thing about this deal is that ARM limited will now provide Hard IP for Intel's process technology.
To understand the importance of this, you have to understand a little more about silicon design and manufacturing than the average Slashdotter. Suppose you are some random fab-less chip designer that builds semi-customized ARM SoCs, a company like Rockchip or Mediatek for example. Generally the way you put together your new SoC is you buy a license for the ARM CPU design, then you buy a license for a GPU design from someone else, then you license a USB controller... so on and so forth, until you have all the building blocks necessary to make your new chip. Then you plug them all together, simulate, fab, validate, and ship.
Those blocks come in two different forms, Hard IP and Soft IP. Soft IP is basically a netlist... its a big text file that lists every transistor in the design and the interconnections between every transistor in the design. Usually soft IP vendors will give you the RTL, which is a more human readable language like Verilog which you compile in to a netlist. Hard IP on the other hard, is more like a vector graphics drawing or a stencil. Hard IP lists every transistor, its x/y coordinates on the silicon, and the exact shape and route of the copper wires. The problem with hard IP is every silicon manufacturer uses different shapes and sizes for their transistors and connecting wires (this is called the process design for the foundry), so a given hard IP design can only be built by the foundry it was designed for.
There is a program called a synthesizer that takes the netlist from the soft IP and generates the layout for the hard IP given a bunch of input parameters that describe the target foundry's process design, rather incredible really. The problem is not every design is "fully synthesizable" for example anything involving high speed I/O or analog (aka the "PHY" layers for modern busses: PCIe, USB, eMMC, Ethernet, SATA etc.) In any case, the pieces of the design that can not be synthesized need to be drawn by hand (aka human hands) using CAD software. For things like CPUs, usually there are some critical pieces that are drawn by hand, because a good human engineer can design a better, more efficient layout than the synthesizer can, at much greater expense of course. So depending on what percentage of your design is not synthesized, switching from one foundry to another can turn out to be a lot of work! This is the important thing here, ARM is providing ready to go hard IP for Intel foundry, just like they do with TSMC already, so the technical barrier for an ARM SoC designer to use Intel foundry is now lower... potentially comparable to TSMC.
Depending on the amount of engineers you have and how sophisticated they are, you might design some of those blocks yourself. Up to the point of companies like Apple and Qualcomm where even the ARM CPU design is a custom implementation and doesn't bear much resemblance to the reference design from ARM limited.
For Intel, using Intel foundry is a non-issue since they have an army of engineers that for the most part they design every IP block themselves anyway. For companies like Apple and Qualcomm that also have armies of engineers switching to Intel foundry is not a technical issue, its about business decisions for them. The big news is the smaller companies that don't have as many resources to do custom design now have Intel foundry as a viable option.
Perhaps the healing and rebuilding of tissues takes nutrients? Also, if you had large open wounds I'm sure your immune system was activated and running on high alert, which also consumes a surprising amount of extra nutrients.
Looks like MacOS and Linux share has remained roughly flat over the last year. Win8.1 use has declined 48.5% and Win7 by 23.1%. Hence Win10's adoption has been at the expense of Win8.1 and to a lesser extent Win7. Overall it seems Microsoft's free upgrade has largely been successful at retaining existing Windows users, but it hasn't won any converts from Apple, and it hasn't slowed down Android at all. They stopped the bleeding, but its not exactly the "threshold" that would return Windows to growth that Microsoft's upper management claimed it would be.
I should have been more clear, every scan *job* not every file. I am very aware of the way Windows works and its horrible amount of overhead for creating a new process compared to UNIX.
This performance concern would need to be balanced against the added security of preventing a persistent malware infection of the scan engine process. Maybe replace the process every 1000 files scanned? This is what performance profiling is for.
From a software engineering standpoint, one should start out with the "best" design and ignore performance for the most part, and then after the initial implementation do performance profiling to see what you actually need to optimize. If you optimize up front, then you have no way of knowing if the optimizations you did in the initial design actually had any benefit.
Since the whole point of it is security it really makes sense to have two copies of your scan engine installed, one in Ring 0 for early boot rootkit detection that scans every driver as it loads and only scans if the binary passes MSFT's driver signing checks first.
All of your scanning of code modules after the kernel is up should be forwarded to a sandboxed user mode service so that even if the scan engine is compromised the malicious code can't go anywhere. Not a bad idea to fire up a new process for every scan so the exploit will be short lived.
Its pretty clear that antivirus software isn't written this way. They run everything in high privileges.
Seriously why the hell does Antivirus software need to run its scan engine at Admin group privileges, and why is half of the scan engine running in Ring 0 kernel drivers?
Its amazing, my work laptop BSODs about once a day just because of some crappy driver included in the Antivirus software installed by IT.
Since it crashes that frequently just in normal operation it seems likely that there is at least 1 vulnerability in that driver which is exploitable from user mode.
Would it have killed them to make it backwards compatible with the hardware that already exists?
Speaking from a purely technical standpoint, based on the way the eMMC and UFS standards are written that would be extremely difficult to achieve. The UFS standard actually uses a MIPIM-PHY design for the actual electrical conveyance of data across the copper data lines. The protocol layer of UFS is actually identical to NVMe and the OS storage drivers interact with the UFS device as if it were a NVMe SSD. By comparison, SD cards have their own proprietary bus format that is derived from MultiMediaCard, which was derived from the SPI bus protocol, which was derived from I2C. This is a completely different hardware and software stack from UFS.
Really what it comes down to is when the UFS specification was originally written it was intended to be an internal bus for giving smartphones faster internal flash. It was not intended to become and external card format that would compete with SD. If that was a consideration from the start, I'm sure JEDEC would have baked a good backwards compatibility story in to the standard. Now that the standard already exists UFS v2.0 needs to be backwards compatible with UFS v1.0, so it is too late to add SD bus compatibility since v1.0 already exists in the market and backwards compatibility with it must be maintained.
Maybe they could try to bake in some SD compatibility without breaking the ability of new cards to work with old UFS hosts after the fact... but given how orthogonal the two designs are that would likely add an unacceptable amount of complexity to the flash chip's controller (remember complexity == more transistors == more expensive controller and more power consumed.)
Without new processors, there's little reason to update the Surface 3 line
I guess whoever wrote this article didn't do very much research. Intel is expected to release Apollo Lake (Cherry Trail's successor) in the 2nd half of 2016.
Slashdot Mirror
I'm not quite sure what the guy that wrote this article is smoking, but there is pretty much no way LinkedIn is going to push MSFT above the $1 trillion mark. Sure the only thing he claims at that MSFT will make $1 trillion before Apple or Alphabet, but at current growth rates none of those companies are going to be in the ball park of $1 trillion valuation for 5-10 years at least, probably longer. It seems unlikely that LinkedIn is the killer acquisition that is going to drive their growth for the next decade (as opposed to all the license revenue they make from existing established business.)
MSFT's current market value is $491.71 billion. Hypothetically, lets say that being acquired by MSFT does not change the value of LinkedIn at all, if that casepost-merger MSFT would be worth... $491.71 billion, they paid $26 billion in stock + cash for LinkedIn reducing the value of MSFT by $26 billion, then the value of LinkedIn ($26 billion) gets added to their market cap, cancelling out the two effects. Now lets say they double the value of LinkedIn by giving it access to their network of enterprise customers, they MSFT's value becomes $491.71 + $26 = $517.71 billion... a 5% increase in the value of MSFT.
I'm sorry, there is just no plausible scenario where LinkedIn suddenly becomes worth more than the entire rest of MSFT is worth. There is no way LinkedIn is going to double the amount of license revenue they generate. That's the problem that enormous businesses like MSFT have... for almost any other business an extra $500 million of quarterly revenue would be an enormous new windfall that would double their stock's value. For MSFT... it would be an extra 2% growth of quarter revenue, to which Wall Street would yawn and say "Here's your gold star MSFT, stock is up 2% after quarterly earnings." Its hard to create big new percentage growth when all the new technologies are not worth anywhere close to your existing massive business. MSFT's partner Intel has also been facing this problem for over a decade. Just like any other company, Apple and Google are not immune to this and are starting to hit the wall that the burden of massive valuation begets. Any company in this position more or less is forced in to paying dividends and becoming a boring "blue chip" stock. Nothing wrong with it really, take the example of 3M, a diversified technology company that has been a stable steady company. They have increased dividend rates every year for over a century now, and their stock holds strong value long term. I think the tech industry as a whole needs to wake up to the fact that we are rapidly turning in to that type of old but strong and steady business.
I'm pretty sure most of the problem is Adobe somehow managed to make the newest version of Flash even slower and more resource consuming than Flash from a year ago. But you absolutely have to make sure you have the newest version installed at all times, since Flash is also a security mess, despite the fact that Adobe has also made keeping Flash up to date a huge PITA (to the point that MSFT and Google handle the updates for them on their respective browsers.) The performance and security of Flash has been a continuing problem for decades now.
Why they bother to continue to develop and release new features for it puzzles me, plugin based content is very clearly a dying platform at this point and its partially Adobe's fault (we wouldn't want to get rid of Flash if it didn't suck so hard.)
Tokamak, sorry autocorrect
I'm sick of these articles that sound like they are mansplaining the basics of tomahawk fusion that we have known since the 1970s and then claims its a new thing. Moreover, they supposedly have a working commercial reactor when we know that a commercial reactor would need to be ITER sized for positive energy generation. Can we keep this crap off slashdot?
The USB power delivery spec standardizes how to increase Vbus voltage and max current. Power profile 5 increases the voltage to 20V with 5A current for 100W of power. It can be implemented on either type A, micro USB, or type C. It used to be that the USB spec only standardized up to 7.5W power draw, which became a limiter on charging time. Now that we have the new power delivery spec extension, there are zero good reasons to implement proprietary charging standards to move beyond 7.5W.
Despite this, Qualcomm is still heavily marketing its proprietary quickcharge 3.0 system to smartphone manufacturers, purely for the incremental profits on licensed wall chargers. I'm glad to see Google throwing their weight around a little in an effort to shut down what is purely a money grab.
It looks like Intel is still playing catch up in the modem space. Interestingly, it looks like for the 2016 iPhone, Apple is using either the Intel XMM 7360 or the Qualcomm X12. Both of these modems were released in 2015. Qualcomm hasn't shipped a new generation since then, but Intel did release the XMM 7480 in February. It would be interesting to see how much progress Intel has made in a year.
Either way, the fact that Intel's modem exists is good for everyone... except Qualcomm. Without it, Qualcomm would be the only LTE modem supplier. There is no doubt Apple is aware of the Intel modem's shortcomings. My guess is Apple is willing to turn a blind eye to that for the "1st gen" product and use the new revenue as a carrot to get Intel to direct its engineering efforts to the features that Apple wants, probably stuff like CDMA for example.
Unless Apple is doing some sort of quasi RAID like read/write access against all available memory chips
"Quasi RAID" is actually how all SSDs work. The controller spreads the writes across multiple flash chips. This is why NVMe is so much faster, the OS can give the SSD controller thousands of outstanding IO requests instead of the max of 32 with SATA. The iPhone uses a single eMMC flash chip which integrates the controller and the NAND on one die. The eMMC chip will do the same thing, only across flash cells instead of entire chips.
In the end, all this comes down to is the fancy 128/256GB eMMC flash comes with a nicer onboard controller than the 32GB one, hence higher bandwidth. That said, with such a huge difference there is no doubt Apple ordered the cheapest 32GB flash they could find. You probably can find the same eMMC chips in a $50 cheap Chinese Android phone. For a $500+ phone they should be paying the extra $2 for higher bandwidth 32GB eMMC chips.
Remember this is Russia. Most of the USSR state owned industry was sold to former Communist party members and other politically well connected people for pennies on the dollar, including the Soyuz factories. It takes 2 years to build a Soyuz because they use the exact same industrial process to build them that they used 40 years ago under Communist rule. It's Govornment subsidized and NASA pays through the teeth for it anyway, so zero incentive to improve inventory management or process efficiency.
This Gartner report seems to tell a very different story than Intel's announcement that Q3 revenue would $700 million better than expected. To quote Intel's press release:
The increase in revenue is primarily driven by replenishment of PC supply chain inventory. The company is also seeing some signs of improving PC demand.
Intel can get in big legal trouble with the SEC for lying on these type of financial announcements, so I tend to give a little more cred to regulated stuff like this than just some random analyst.
Now before anyone freaks out, keep in mind that there is nothing in the Audio over USB-C spec that requires the device that implements it to not have a headphone jack. Its totally legit for a phone to support this spec and have a headphone jack. In fact, I suspect that most vendors will probably go this route.
Being able to plug your phone in to a single USB-C connector on your car or stereo and have it charge the battery and play music using 1 wire is a nice feature.
Really should read "UK's Top Police Warn That Making Aim-Bots/Game Cheats May Turn Kids into Cyber Criminals"
I'm not an expert in sociology, but it seems plausible that unethical behavior in online video games can be a gateway to unethical online behavior in general. From a technical standpoint I know that the skills developed by hacking games are similar to the skills needed to hack financial software.
I don't think this made the Slashdot front page but Intel bought VIA's CDMA modem design and license about a year ago. Intel's modems currently only support GSM & LTE, whereas VIA never updated their CDMA modems to be LTE capable. It likely will take a couple years for Intel to integrate VIA's CDMA implementation with their LTE design, but once its done, Intel's modems will be just as capable as Qualcomm's.
CMDA isn't only important for the US Verizon/Sprint market, the much more important reason to implement CDMA is China Telecom. Either way, the iPhone 7S will likely mark the return of all iPhones being universally supported by all carriers... regardless of whether there is an Intel or Qualcomm modem inside it.
No word from Apple about whats going to happen with the Apple Watch Edition. Seriously if you are going to sell a $17,000 luxury item that will be obsolete 16 months after it is released you better have a good upgrade plan in place.
Its a little disingenuous to say that Watson "created" the trailer. The only thing Watson did was run a pattern recognition algorithm to figure out which clips in the movie were tense, happy, scary, etc. Then a human editor sorted through all of the clips, picked the good ones and put them in sequence to create a trailer that actually had narrative instead of just being a hodge podge of disjoint clips.
Pattern recognition is getting better which is the first step to creating an AI... but Watson, and AI in general is still very far off from creating a computer program that is capable of original thought.
Java is now toxic thanks to its owner. For the sake of the entire tech industry, we all should consider it a legacy technology that should be removed from everything as quickly as possible. Unfortunately that will take years... maybe even decades, but we must start the deprecation process as quickly as possible. Besides, in the 20 years since it was created we have better cross platform languages now anyway.
Thankfully a lot of us have input in to technical decisions here. We all need to take a stand and kill Java.
I keep a Windows system around for minor software that needs it
AKA "games".
Other than games, the very important thing that keeps Windows on my personal system is TurboTax. Like pretty much any other US Taxpayer that has a tax situation too complex for form 1040-EZ and doesn't want to pay ~$150 for H&R Block or ~$300 for a certified CPA. I hired a CPA once and $50 per year TurboTax did a better job!
Before anyone says Wine, its a non-starter. TurboTax uses a bunch of .Net features that don't work 100% right on Wine like WPF. Unfortunately the Mac version is absolute garbage, so that route isn't viable either. It really sucks, but the easiest way to be a lawful US citizen is to have a Windows system.
Trying to argue that Chrome isn't trans-formative when compared to Java SE is so ridiculous its almost funny. You can only run litigious business like Oracle based on making everyone else pay a protection racket for so long.
The fact that Intel is offering to manufacture ARM cores for their custom foundry customers is not new. In fact, there are some Altera FPGAs with embedded ARM cores being manufactured by Intel already. The important thing about this deal is that ARM limited will now provide Hard IP for Intel's process technology.
To understand the importance of this, you have to understand a little more about silicon design and manufacturing than the average Slashdotter. Suppose you are some random fab-less chip designer that builds semi-customized ARM SoCs, a company like Rockchip or Mediatek for example. Generally the way you put together your new SoC is you buy a license for the ARM CPU design, then you buy a license for a GPU design from someone else, then you license a USB controller... so on and so forth, until you have all the building blocks necessary to make your new chip. Then you plug them all together, simulate, fab, validate, and ship.
Those blocks come in two different forms, Hard IP and Soft IP. Soft IP is basically a netlist... its a big text file that lists every transistor in the design and the interconnections between every transistor in the design. Usually soft IP vendors will give you the RTL, which is a more human readable language like Verilog which you compile in to a netlist. Hard IP on the other hard, is more like a vector graphics drawing or a stencil. Hard IP lists every transistor, its x/y coordinates on the silicon, and the exact shape and route of the copper wires. The problem with hard IP is every silicon manufacturer uses different shapes and sizes for their transistors and connecting wires (this is called the process design for the foundry), so a given hard IP design can only be built by the foundry it was designed for.
There is a program called a synthesizer that takes the netlist from the soft IP and generates the layout for the hard IP given a bunch of input parameters that describe the target foundry's process design, rather incredible really. The problem is not every design is "fully synthesizable" for example anything involving high speed I/O or analog (aka the "PHY" layers for modern busses: PCIe, USB, eMMC, Ethernet, SATA etc.) In any case, the pieces of the design that can not be synthesized need to be drawn by hand (aka human hands) using CAD software. For things like CPUs, usually there are some critical pieces that are drawn by hand, because a good human engineer can design a better, more efficient layout than the synthesizer can, at much greater expense of course. So depending on what percentage of your design is not synthesized, switching from one foundry to another can turn out to be a lot of work! This is the important thing here, ARM is providing ready to go hard IP for Intel foundry, just like they do with TSMC already, so the technical barrier for an ARM SoC designer to use Intel foundry is now lower... potentially comparable to TSMC.
Depending on the amount of engineers you have and how sophisticated they are, you might design some of those blocks yourself. Up to the point of companies like Apple and Qualcomm where even the ARM CPU design is a custom implementation and doesn't bear much resemblance to the reference design from ARM limited.
For Intel, using Intel foundry is a non-issue since they have an army of engineers that for the most part they design every IP block themselves anyway. For companies like Apple and Qualcomm that also have armies of engineers switching to Intel foundry is not a technical issue, its about business decisions for them. The big news is the smaller companies that don't have as many resources to do custom design now have Intel foundry as a viable option.
Perhaps the healing and rebuilding of tissues takes nutrients? Also, if you had large open wounds I'm sure your immune system was activated and running on high alert, which also consumes a surprising amount of extra nutrients.
The data over at Stat Counter seems to agree:
http://gs.statcounter.com/#desktop-os-ww-monthly-201506-201606
Looks like MacOS and Linux share has remained roughly flat over the last year. Win8.1 use has declined 48.5% and Win7 by 23.1%. Hence Win10's adoption has been at the expense of Win8.1 and to a lesser extent Win7. Overall it seems Microsoft's free upgrade has largely been successful at retaining existing Windows users, but it hasn't won any converts from Apple, and it hasn't slowed down Android at all. They stopped the bleeding, but its not exactly the "threshold" that would return Windows to growth that Microsoft's upper management claimed it would be.
I should have been more clear, every scan *job* not every file. I am very aware of the way Windows works and its horrible amount of overhead for creating a new process compared to UNIX.
This performance concern would need to be balanced against the added security of preventing a persistent malware infection of the scan engine process. Maybe replace the process every 1000 files scanned? This is what performance profiling is for.
From a software engineering standpoint, one should start out with the "best" design and ignore performance for the most part, and then after the initial implementation do performance profiling to see what you actually need to optimize. If you optimize up front, then you have no way of knowing if the optimizations you did in the initial design actually had any benefit.
Since the whole point of it is security it really makes sense to have two copies of your scan engine installed, one in Ring 0 for early boot rootkit detection that scans every driver as it loads and only scans if the binary passes MSFT's driver signing checks first.
All of your scanning of code modules after the kernel is up should be forwarded to a sandboxed user mode service so that even if the scan engine is compromised the malicious code can't go anywhere. Not a bad idea to fire up a new process for every scan so the exploit will be short lived.
Its pretty clear that antivirus software isn't written this way. They run everything in high privileges.
Seriously why the hell does Antivirus software need to run its scan engine at Admin group privileges, and why is half of the scan engine running in Ring 0 kernel drivers?
Its amazing, my work laptop BSODs about once a day just because of some crappy driver included in the Antivirus software installed by IT.
Since it crashes that frequently just in normal operation it seems likely that there is at least 1 vulnerability in that driver which is exploitable from user mode.
Would it have killed them to make it backwards compatible with the hardware that already exists?
Speaking from a purely technical standpoint, based on the way the eMMC and UFS standards are written that would be extremely difficult to achieve. The UFS standard actually uses a MIPI M-PHY design for the actual electrical conveyance of data across the copper data lines. The protocol layer of UFS is actually identical to NVMe and the OS storage drivers interact with the UFS device as if it were a NVMe SSD. By comparison, SD cards have their own proprietary bus format that is derived from MultiMediaCard, which was derived from the SPI bus protocol, which was derived from I2C. This is a completely different hardware and software stack from UFS.
Really what it comes down to is when the UFS specification was originally written it was intended to be an internal bus for giving smartphones faster internal flash. It was not intended to become and external card format that would compete with SD. If that was a consideration from the start, I'm sure JEDEC would have baked a good backwards compatibility story in to the standard. Now that the standard already exists UFS v2.0 needs to be backwards compatible with UFS v1.0, so it is too late to add SD bus compatibility since v1.0 already exists in the market and backwards compatibility with it must be maintained.
Maybe they could try to bake in some SD compatibility without breaking the ability of new cards to work with old UFS hosts after the fact... but given how orthogonal the two designs are that would likely add an unacceptable amount of complexity to the flash chip's controller (remember complexity == more transistors == more expensive controller and more power consumed.)
Without new processors, there's little reason to update the Surface 3 line
I guess whoever wrote this article didn't do very much research. Intel is expected to release Apollo Lake (Cherry Trail's successor) in the 2nd half of 2016.