or Connect
AppleInsider › Forums › Mobile › iPad › Apple GPU vendor Imagination buys MIPS, ARM licenses patents
New Posts  All Forums:Forum Nav:

Apple GPU vendor Imagination buys MIPS, ARM licenses patents

post #1 of 23
Thread Starter 
Imagination Technologies, the source of PowerVR GPU cores used in Apple's A-series iOS chips, has acquired MIPS, one of the few remaining CPU architectures outside of Intel, ARM and PowerPC.

According to a report by Anandtech Imagination will buy MIPS for $60 million in cash, a figure that includes 82 patents related to the firm's MIPS processor architecture.

Another 580 patents and patent applications owned by MIPS were sold for $350 million to Bridge Crossing, a consortium that includes ARM and other members of the Allied Security Trust, including HP, IBM and Intel. The consortium works to prevent critical CPU patents from being obtained by patent trolls.

Apple secretly licensed advanced PowerVR GPU technology from Imagination in 2008, back when the company was expected to be using Intel's Silverthorne (now called Atom) chips for its anticipated new tablet.

Rather than going with Intel, Apple developed its own custom chip design based on ARM CPU cores and PowerVR graphics cores, a design the company called the A4, and which it subsequently used in the original iPad and iPhone 4.

Apple purchased a 3 percent stake in Imagination with 8 million shares, and continues to use the company's advanced graphics in its latest chip designs, including the three and four core A6 and A6X, used in the iPhone 5 and new iPad 4.

Apple's A-series chips


The MIPS architecture was introduced in the early 1980s as a new RISC design, and greatly influenced subsequent architectures including DEC's Alpha. Originally used in workstations, SGI's MIPS chips were targeted by Microsoft's failed mid-90s efforts to port Windows NT across CPU architectures.

MIPS ended up being widely used in embedded devices, including early Windows CE mobile products and Sony's PlayStation 2 and PSP and Nintendo's N64 game console.

The elimination of MIPS leaves ARM and Intel with increasingly fewer competitors among CPU architectures. However, the open licensing of ARM's IP has resulted in extensive competition among its licensees, with Apple being a leading producer and consumer of ARM-based chips.

Even among ARM licensees however, competition among high end chip designs is dwindling. Texas Instruments recently announced plans to exit the mobile chip business, killing off the OMAP-5 ARM design that would have directly competed with Apple's A6.
post #2 of 23

The evidence is beginning to pile up. 

post #3 of 23
Quote:
Originally Posted by AppleInsider View Post

The elimination of MIPS leaves ARM and Intel with increasingly fewer competitors among CPU architectures.

How does the acquisition of MIPS by Imagination equate to the elimination of MIPS?
post #4 of 23
Quote:
Originally Posted by Quadra 610 View Post

The evidence is beginning to pile up. 

Evidence of what?
post #5 of 23
Summary : Imagination bought 82 patents for $60 million in cash and ARM, HP, IBM, Intel consortium bought 580 patents for $350 million.
post #6 of 23
MIPS chips were targeted by Microsoft's failed mid-90s efforts to port Windows NT across CPU architectures.

 

And that failure is the major reason why Microsoft and Intel are trailing so badly in mobile right now.  If Microsoft had successfully ported the NT kernel to RISC (or if they had allowed that porting project to succeed) then they would have been able to migrate it to ARM for mobile devices.  The way Apple migrated OS X from PowerPC to Intel and later optimize it for ARM and call it iOS.  

 

But no, the powers-that-be at Microsoft went with Intel for various reasons.  And for 20 years, neither company saw any need for optimization.  Microsoft let Windows get more and more bloated because Moore's law meant that CPU speeds would always increase to compensate for the bloat.  And desktop computers would always be plugged in, so there was no need for Intel to work on power efficiency.  Neither company bothered doing much optimization.  And now Microsoft and Intel are paying the price in mobile, where software and hardware optimization are crucial.


Edited by SockRolid - 11/6/12 at 10:59am

Sent from my iPhone Simulator

Reply

Sent from my iPhone Simulator

Reply
post #7 of 23
Quote:
Originally Posted by SockRolid View Post

And that failure is the major reason why Microsoft and Intel are trailing so badly in mobile right now.  If Microsoft had successfully ported the NT kernel to RISC (or if they had allowed that porting project to succeed) then they would have been able to migrate it to ARM for mobile devices.  The way Apple migrated OS X from PowerPC to Intel and later optimize it for ARM and call it iOS.  

But no, the powers-that-be at Microsoft went with Intel for various reasons.  And for 20 years, neither company saw any need for optimization.  Microsoft let Windows get more and more bloated because Moore's law meant that CPU speeds would always increase to compensate for the bloat.  And desktop computers would always be plugged in, so there was no need for Intel to work on power efficiency.  Neither company bothered doing much optimization.  And now Microsoft and Intel are paying the price in mobile, where software and hardware optimization are crucial.

Is the success being referred to a working OS on different architecture, or market uptake? The ports were functional, and there was some hardware, but the market really didn't seem to care much. There were fully functional PPC, MIPS and Alpha ports of Windows NT 4 on the installer CD. There were even beta versions of Windows 2000 for the Alpha platform. I had used Windows NT 4 on Alpha for about five years.

I don't know if your characterization on the mobile impact was necessarily right, notebook computers were starting to be popular around that time. I do agree that power optimization was not taken very seriously in the Wintel camp.
Edited by JeffDM - 11/6/12 at 11:17am
post #8 of 23
Quote:
Originally Posted by SockRolid View Post

 

And that failure is the major reason why Microsoft and Intel are trailing so badly in mobile right now.  If Microsoft had successfully ported the NT kernel to RISC (or if they had allowed that porting project to succeed) then they would have been able to migrate it to ARM for mobile devices.  The way Apple migrated OS X from PowerPC to Intel and later optimize it for ARM and call it iOS.  

 

But no, the powers-that-be at Microsoft went with Intel for various reasons.  And for 20 years, neither company saw any need for optimization.  Microsoft let Windows get more and more bloated because Moore's law meant that CPU speeds would always increase to compensate for the bloat.  And desktop computers would always be plugged in, so there was no need for Intel to work on power efficiency.  Neither company bothered doing much optimization.  And now Microsoft and Intel are paying the price in mobile, where software and hardware optimization are crucial.

NT supported 4 different CPU architectures. windows 2000 dropped support because almost no one used them. they were too expensive compared to Intel

post #9 of 23

Apple purchased a 3 percent stake in Imagination with 8 million shares. Why 3% ?

Citing unnamed sources with limited but direct knowledge of the rumoured device - Comedy Insider (Feb 2014)
Reply
Citing unnamed sources with limited but direct knowledge of the rumoured device - Comedy Insider (Feb 2014)
Reply
post #10 of 23

DED is mistaken -- TI is not killing OMAP5

post #11 of 23
Quote:
Originally Posted by Ireland View Post

Apple purchased a 3 percent stake in Imagination with 8 million shares.
 Why 3% ?

1) They probably what was available and what they thought was needed to get a seat on the board or some sort of other input.

2) They appear to hold to about 9.5%.

This bot has been removed from circulation due to a malfunctioning morality chip.

Reply

This bot has been removed from circulation due to a malfunctioning morality chip.

Reply
post #12 of 23
Quote:
Originally Posted by Ireland View Post

Apple purchased a 3 percent stake in Imagination with 8 million shares. Why 3% ?

 

Because they liked that number, maybe

post #13 of 23
Quote:
Originally Posted by Ireland View Post

Apple purchased a 3 percent stake in Imagination with 8 million shares. Why 3% ?

 

Time for Apple to up the stakes! Also, shouldn't they have initiated their stock buyback by now? I think their own stock is a relative bargain right now.

Proud AAPL stock owner.

 

GOA

Reply

Proud AAPL stock owner.

 

GOA

Reply
post #14 of 23
Back in those days everything was far slower so MIPS was not up to the job of running Windows. These days it is very different. And in fact it was only really Apple's iPad 1 that was not up to the task of running Flash video.
post #15 of 23
Quote:
Originally Posted by aBeliefSystem View Post

Back in those days everything was far slower so MIPS was not up to the job of running Windows. These days it is very different. And in fact it was only really Apple's iPad 1 that was not up to the task of running Flash video.

Nonsense.

For many tasks, MIPS was faster than x86 at the time.

And it's not just the iPad 1 that couldn't run Flash. Even Adobe admitted that there were no mobile platforms that ran it well.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #16 of 23
Quote:
Originally Posted by jragosta View Post


Nonsense.
For many tasks, MIPS was faster than x86 at the time.
And it's not just the iPad 1 that couldn't run Flash. Even Adobe admitted that there were no mobile platforms that ran it well.

 

It was all about cost and the market conditions at the time.  MIPS were faster than Intel chips when they first came out, but clock speeds were advancing quickly, so any advantages MIPS had with a more efficient RISC design, Intel quickly overcame by cranking out faster x86 chips.  Intel had a huge market share and fabrication facilities - MIPS did not.   MIPS may have had a theoretically better CPU design, but Intel was way ahead in developing chip manufacturing processes.

 

Even though Microsoft did release NT with support for MIPS, even x86 NT did not do great business for the first year or two.  Most business software was either still DOS or Windows 3.1.  Developers coded all kinds of tricks to get the most out of 640MB of RAM.  Win 3.1 apps did not just run on Windows NT,  lots of it had to be rewritten.   So there was not much software to run on the MIPS NT computers.

 

MIPS had essentially become an intellectual property company a long time ago and got out of the chip making business, but sold their designs and IP to others to use.  MIPS processors are all over - but embedded or in some device besides a PC.  And some of the work MIPS was founded on was ground breaking stuff, so their patents are probably very significant.  

 

Intergraph was once similar to MIPS - they had their own CPU design called Clipper and focused on the design/CAD workstation market.  They ended up getting out of the CPU business and focusing on software, but they made huge money suing Intel over the use of technology covered by Intergraph CPU patents - this was long before everyone was suing everyone else.

 

Smart buy by Imagination.  They can continue to license to MIPS IP to other chip builders and they are a lot stronger in their patent portfolio.

post #17 of 23
Quote:
Originally Posted by SockRolid View Post

 

And that failure is the major reason why Microsoft and Intel are trailing so badly in mobile right now.  If Microsoft had successfully ported the NT kernel to RISC (or if they had allowed that porting project to succeed) then they would have been able to migrate it to ARM for mobile devices.  The way Apple migrated OS X from PowerPC to Intel and later optimize it for ARM and call it iOS.  

 

But no, the powers-that-be at Microsoft went with Intel for various reasons.  And for 20 years, neither company saw any need for optimization.  Microsoft let Windows get more and more bloated because Moore's law meant that CPU speeds would always increase to compensate for the bloat.  And desktop computers would always be plugged in, so there was no need for Intel to work on power efficiency.  Neither company bothered doing much optimization.  And now Microsoft and Intel are paying the price in mobile, where software and hardware optimization are crucial.

 

I think what is missing here is that what Apple accomplished (well NeXT) was non-trivial. They were much more nimble at keeping their OS abstracted from the CPU -- and they accomplished the transition from PowerPC to Intel with barely a hiccup. I've still got some old PowerPC devices around, and since then, Apple has transitioned elegantly via Carbon apps.

 

They also had to deal with "big endian" and "little endian" math reversals. Really, they had to speak two different languages for some time, and then do it again with mobile computing. They made it look easy. 

 

The CRUCIAL DIFFERENCE however between Microsoft and Apple, is that the latter is good at "obsoleting itself." Microsoft languishes with petty fiefdoms that see each other as a greater threat than any other company, and their OS has 5 different ways to do the same thing (usually buried in an "advanced" tab) and they are most of them are mediocre and the result is non-intuitive. You just have to KNOW the right way to do things.

 

Apple keeps abstracting, and creating objects. And they keep breaking things that work to make something new that's simpler and more elegant. They are the first to "cast off" things that no longer make sense, without abandoning compatibility.

 

I think it's pretty awesome. They've avoided the kind of megalith syndrome that seems to affect large old companies who wind up ossifying on whatever they can monopolize, and making sure that customers stay with that, rather than focusing on what the customer will need in the future.

 

I didn't want that to be "Apple worship" -- it's more just taking a moment and appreciating a company "getting it right." Because we have to recognize what Microsoft DID RIGHT with their developer tools and platform even while abusing the monopoly so well with licensing.

 

And SPEAKING of monopoly -- I'm wondering if Apple is going to purchase this vendor, or if they have their OWN chip designs waiting in the wings. A third Chip powerhouse is actually necessary in the marketplace, because Intel has so soundly defeated AMD in the marketplace (even after faltering for quite a while). Intel is a ruthless competitor, and if they buy up the IP for most of the mobility platform -- then it's going to be another 10 years before any real competition can enter the marketplace again.

 

I'm not holding any hope that our justice system would prevent any monopoly in the mobile chip market. So I'm guessing that the STRATEGY here, is that Imagination is setting itself up as a "takeover target." They'll hold the critical IP and some company either buys them to control the market, or Google buys them as they did Motorola defensively (or maybe as a patent troll?). And wow -- Google really should have gone after Imagination rather than buy "stale" IP from Motorola. Microsoft is probably SO HAPPY, they didn't complete that acquisition of AOL/Time Warner as well.

 

Why is this stuff so obvious to me and I'm not rich? There's got to be some other factor beyond insight, vision and wisdom that has to do with wealth... ;)

post #18 of 23
Quote:
Originally Posted by aBeliefSystem View Post

Back in those days everything was far slower so MIPS was not up to the job of running Windows. These days it is very different. And in fact it was only really Apple's iPad 1 that was not up to the task of running Flash video.

I am sure that the reason Adobe abandoned Flash was because it ran so well on everything except for Apple's iPad 1.

"A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools." Douglas Adams

Reply

"A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools." Douglas Adams

Reply
post #19 of 23
Quote:
Originally Posted by SpamSandwich View Post

 

Time for Apple to up the stakes! Also, shouldn't they have initiated their stock buyback by now? I think their own stock is a relative bargain right now.

 

That 3% purchase sounds like a way to "keep their finger on the pulse" but not commit. Remember when Microsoft bought Kinect from the developers and then we were freaking "Why didn't Apple do that -- everyone is going to interface with this machine in the future." Now we see that "Leap Technologies" is going to run rings around Kinect. 

 

So really, we don't know what we don't know. However, if we can assume Apple is VERY AWARE of the importance of MIPS and the mobile chip platform -- then it's likely the haven't snapped up Imagination for a reason. They spent more money buying stock in parts manufacturers (like for their screens) than in these chip companies.

 

If Apple does not BUY OUT, Imagination in the near future -- Either there is a MUCH BETTER startup in the mobile space hiding in the wings, or Apple has something new up their sleeve -- I just can't imagine they aren't aware of the risks here of either Intel or Google acquiring critical IP for mobile computers.

post #20 of 23
Quote:
Originally Posted by Fake_William_Shatner View Post


Why is this stuff so obvious to me and I'm not rich? There's got to be some other factor beyond insight, vision and wisdom that has to do with wealth... 1wink.gif

Contemplate your willingness to take big risks, based upon your insight, vision, and wisdom.
post #21 of 23
Quote:
Originally Posted by Fake_William_Shatner View Post

 

Why is this stuff so obvious to me and I'm not rich? There's got to be some other factor beyond insight, vision and wisdom that has to do with wealth... ;)

 

There is, its called luck.

 

J.

post #22 of 23

As a former SGI developer, I loved the MIPS design, and I'd wish to see the MIPS architecture alive again, although I suppose this is an acquisition with other interests rather than the MIPS architecture itself.

post #23 of 23
Quote:
Originally Posted by JeffDM View Post


Is the success being referred to a working OS on different architecture, or market uptake? The ports were functional, and there was some hardware, but the market really didn't seem to care much. ...

 

You are absolutely correct. People seem to forget if they ever knew that the original development systems for the Xbox 360 were Apple Power Mac G5s running Windows NT. This was years after Microsoft "dropped" PPC support in Windows NT. 

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: iPad
AppleInsider › Forums › Mobile › iPad › Apple GPU vendor Imagination buys MIPS, ARM licenses patents