Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: ferrellsl on August 06, 2013, 07:53:06 PM
-
This should have happened about 10 years ago......I'm thinking it's too little, too late, to be of any consequence.
http://arstechnica.com/information-technology/2013/08/ibm-to-begin-arm-style-licensing-of-power-cpu-architecture/
-
Be interesting to see where this goes. PPC is still popular in the embedded market, so manufacturers being able to integrate it into their own SoCs could work out pretty well. But I'll say what I've been saying about ARM for the last five years: somebody give it to me in a laptop form factor already!
-
This should have happened about 10 years ago......I'm thinking it's too little, too late, to be of any consequence.
Ditto. Maybe even 20 years ago tbh.
-
Release a 5 GHz version and some beefed up cache etc and they may have chance in h-ll .. ;)
-
But I'll say what I've been saying about ARM for the last five years: somebody give it to me in a laptop form factor already!
You mean this?
http://www.amazon.com/Samsung-XE303C12-A01US-Chromebook-Wi-Fi-11-6-Inch/dp/B009LL9VDG/ref=sr_1_1?ie=UTF8&qid=1375819751&sr=8-1&keywords=chromebook
-
You mean this?
http://www.amazon.com/Samsung-XE303C12-A01US-Chromebook-Wi-Fi-11-6-Inch/dp/B009LL9VDG/ref=sr_1_1?ie=UTF8&qid=1375819751&sr=8-1&keywords=chromebook
No, that CPU is an ARM derivative. He wants a PPC laptop that can perform as well as the AMD/x86/ARM laptops that are out there.
-
This should have happened about 10 years ago......I'm thinking it's too little, too late, to be of any consequence.
http://arstechnica.com/information-technology/2013/08/ibm-to-begin-arm-style-licensing-of-power-cpu-architecture/
Interesting. But how "new" is this, really?
Did PA Semi not get a license?
Do AMCC not license it?
Did Xilinx not license it?
Nintendo, Sony, Microsoft?
I'm not sure how this new thing differs from the old thing.
-
Interesting. But how "new" is this, really?
Did PA Semi not get a license?
Do AMCC not license it?
Did Xilinx not license it?
Nintendo, Sony, Microsoft?
I'm not sure how this new thing differs from the old thing.
This differs from the "old thing" in a few ways.
The first lot got architecture licenses - they had to then design their own implementation from the ground up and source manufacturing, then go through all of the pain & expense of bringing a design to fabrication.
Nintendo, Sony and Microsoft paid IBM to design an implementation for them - or actually to customise one they already had in some cases and then fabricate it for them.
This will be the first time that IBM license the design of an actual implementation that can then be further customised by the licensee and fabricated wherever they then like. That means with whatever GPU they like, bus, cache sizes etc. All built atop of a design that's already known good and can be implemented at less cost.
It's one of the big differences between things like ARM and traditional CPUs and is credited with being one of the reasons that ARM have had such success compared to them. They're much more flexible than the traditional chip houses that seem to want to decide how, when and why you get to use their designs. ARM will let you do pretty much whatever you like.
I doubt IBM will go the whole hog and only start charging you for the chips you actually make like ARM do, but you never know.
-
Too little, too late to matter on a big enough scale.
-
I really loved the CELL
so genius.
-
Maybe the Chinese will pick up on this and develop some interesting new PPC driven systems.
-
With Sony & Xbox both dropping PPC for next gen consoles IBM is probably looking for more PPC customers.
-
It very interesting. I wish IBM success. Might be little biased. I have a family member that works for IBM.
-
Maybe the Chinese will pick up on this and develop some interesting new PPC driven systems.
Why when they can have dirt cheap ARM in volume and not have to reinvent the wheel? I can't see the embedded network folks forking over that type of risky investment to do R&D and then production when Freescale is producing PPCs that are capable of doing the job.
-
Licensing will not help Amiga. Unless someone is going to do a large batch of G4's and G5's (Will it now be possible?) there won't be any advantage.
We don't need people throwing more spanners in the works, I say go ARM and stick with that course.
-
Licensing will not help Amiga. Unless someone is going to do a large batch of G4's and G5's (Will it now be possible?) there won't be any advantage.
We don't need people throwing more spanners in the works, I say go ARM and stick with that course.
Why should someone create a big bunch of G4 and G5? For which customers? (Amiga-market is too small to be of any interest)
-
Too little, too late to matter on a big enough scale.
Depends on your goals. The PPC market is still larger than the x86 market in number of units shipped. Something like ~500 million PPC cpu's are shipped a year - comparable to MIPS.
Last I checked x86 was closer to 300 million units.
POWER/PPC is clearly still viable as an architecture. The question is if IBM licensing their high end designs can make it viable for market segments like cell phones, tablets or servers (desktops are "hopeless" as long as Windows is so entrenched) where it isn't really present.
(And if WDC's numbers are right, 6502 cores might outstrip x86 too...)
Of course ARM dwarfs all of them, with an estimated 3 billion cpu's likely to get shipped this year...
-
" The PPC market is still larger than the x86 market in number of units shipped"
Really?
-
" The PPC market is still larger than the x86 market in number of units shipped"
Really?
They did alright with the current games console generation as PPC was used by all three. Next year they will likely take a beating.
-
They did alright with the current games console generation as PPC was used by all three. Next year they will likely take a beating.
I "googled" a little. Most reports were older (about up to 2005). Newer reports are only about Xbox using X86 in future and similar (hardly good news). But for "Amiga-community" it does not change much. That "AmigaOS" or "MorphOS" using PPC become competitive again hardly anyone expects. If IBM really drops (to a wide scale) PPC development only would make clear again that PPC is a dead end, a small niche for some geeks at best. To get outside a change in architecture is needed (be it ARM or X86/X64). But most are happy with their exclusive hobby and the situation so finally it is not important.
-
The PowerPC embedded market is still quite large, don't forget that - and the x86 embedded market much less so. So I'm not surprised at all if PPC outsells x86 overall.
-
might be... I have not found numbers
the problem is "Amiga" (all variants) is designed as desktop and not for embedded market (AmigaOS/MorphOS would need desktop processors). So if at all someone invests money in PPC development it will be for embedded devices (where performance is less important). And if I would invest in embedded devices today I would invest in ARM propably. Processor development is risky and expensive and companies tend to invest in "security" and ARM today is the more secure bet.
-
" The PPC market is still larger than the x86 market in number of units shipped"
Really?
Last time I dug up numbers, the PPC market was estimated at ~500 million a year, comparable to MIPS (and far outstripped by ARM - ARM estimates 3 billion cores/year).
X86 meanwhile were present in ~350-360 million desktops, servers and laptops last year, and embedded usage of x86 is almost a rounding error, so more than ~400 million total is unlikely.
(As an amusing aside, Western Design Centre also still claims "hundreds of millions" of 6502 instruction set cores per year, though a substantial portion of that is likely embedded in custom ASICs)
In terms of *revenue*, though, Intel makes about *7 times* more from their CPUs that Qualcomm (at second place) makes from their CPUs, as a result of a ridiculously higher average revenue per unit.
-
They did alright with the current games console generation as PPC was used by all three. Next year they will likely take a beating.
The combined lifetime sales of the current generation games consoles only adds up to 4-5 months worth of PPC sales - losing them won't make much of a dent in terms of units. It will hurt worse in revenue.
-
Maybe it's just me but the keyword here is not IBM or PowerPC it's NVidia and possibly Google ...
You don't need latest GPU integrated on a PPC SoC to use it in routers ...
Does anyone know how compatible new Power8 is to say G5 ?
-
The combined lifetime sales of the current generation games consoles only adds up to 4-5 months worth of PPC sales - losing them won't make much of a dent in terms of units. It will hurt worse in revenue.
I'd like to see evidence of that, my google searches haven't shown that anyone is shipping more powerpc's than sony/Microsoft/Nintendo are right now.
Freescale, LSI and AppliedMicro are moving to Arm, because PowerPC isn't selling. IBM are still using them in their servers, but those are low number of sales with a high revenue
-
Last time I dug up numbers, the PPC market was estimated at ~500 million a year, comparable to MIPS (and far outstripped by ARM - ARM estimates 3 billion cores/year).
X86 meanwhile were present in ~350-360 million desktops, servers and laptops last year, and embedded usage of x86 is almost a rounding error, so more than ~400 million total is unlikely.
(As an amusing aside, Western Design Centre also still claims "hundreds of millions" of 6502 instruction set cores per year, though a substantial portion of that is likely embedded in custom ASICs)
In terms of *revenue*, though, Intel makes about *7 times* more from their CPUs that Qualcomm (at second place) makes from their CPUs, as a result of a ridiculously higher average revenue per unit.
x86 outsells PowerPC even in the embedded market. Look it up, if you don't believe me. And the trend now is strongly towards x86 and ARM.
-
I'd like to see evidence of that, my google searches haven't shown that anyone is shipping more powerpc's than sony/Microsoft/Nintendo are right now.
Freescale, LSI and AppliedMicro are moving to Arm, because PowerPC isn't selling.
Indeed Freescale are currently producing some great ARM M4 chips!
-
@vidarh
Keep in mind, that the ppc cpus in washingmachines, microwaves o r printers, as well as routers don't count as if it were computers.
-
Is PPC more effiient than ARM etc when counting instructions/herts and instructions/watt?
-
(http://i40.tinypic.com/wcbqfb.jpg)
Since 2011, the market has shifted even more in favour of ARM and x86.
By revenue, for 2012
(http://i40.tinypic.com/116ndw6.png)
Bottomline, PPC is on it's way out.
-
@vidarh
Keep in mind, that the ppc cpus in washingmachines, microwaves o r printers, as well as routers don't count as if it were computers.
The same could be said of the ARM CPU's used in the very same devices.
-
@vidarh
Keep in mind, that the ppc cpus in washingmachines, microwaves o r printers, as well as routers don't count as if it were computers.
I've seen a few PPC chips in some Engine Control Units for cars and also as one of the redundant pathways in some aircraft avionics (where the flight control system will have two separate but functionally identical systems, built using different components I.e. one system would use PPC the other would use an ARM)... But other than that... I've never seen PPC in consumer electronics since the last PowerBook...
And the only reason the PPC is chosen for the aircraft and car systems, is becuase the chips have been built, tested and certified for harsh environments... Nothing to do with any architecture wins for the PPC :(
-
Is PPC more effiient than ARM etc when counting instructions/herts and instructions/watt?
The ARM is available in so many flavours now, there is a model that fits any application bette than any available PPC... The ARM always had the advantage of being competitive even when fabricated on older processes, the PPC was only ever competitive when fabed on modern processes :-/
-
Since 2011, the market has shifted even more in favour of ARM and x86.
I'd question that source. For example, where are Qualcomm and Samsung, the by far two largest (by revenue) ARM licensees on the last chart?
-
@vidarh
Keep in mind, that the ppc cpus in washingmachines, microwaves o r printers, as well as routers don't count as if it were computers.
We're talking about "computers", we're talking about CPU's. By that argument about 90% of ARM's volume is irrelevant as well.
-
For those who are interested, the source of the graph posted was here:
http://chipdesignmag.com/sld/schirrmeister/tag/intel/
I think this has a bearing on things:
The underlying data for this graph includes standalone CPUs and System-on-Chips (SoCs) consumed by all tracked electronic system types.
-
For those who are interested, the source of the graph posted was here:
http://chipdesignmag.com/sld/schirrmeister/tag/intel/
I think this has a bearing on things:
Even then it doesn't make any sense. ST MicroElectronics had about $8.5 billion total revenue for fiscal 2012, and according to their last quarterly report they're on target for similar revenue this year. Qualcomm and Samsung each made about $5 billion on ARM CPU's last year. (EDIT: Notice the difference - both Qualcomm and Samsung are substantially larger than STM if you look at total revenue - this is CPU revenue *only* compared to total revenue for STM)
Never mind that most of STM's CPU's business is in ST Ericsson - a 50/50 joint venture that only accounts for about $300 million of STM's yearly revenues.
STM *may* belong there if they're counting 8-bit microcontrollers, but in that case companies like Zilog should be there too (they bought Samsungs 4-bit and 8-bit microcontroller lines), and architectures like the ST32, Z80, 6502, 8051 and others would also in that case compete in the volume charts. Incidentally, NXP exited the 8051 market at the beginning of last year, and Zilog entered the same market.. (EDIT: Zilog is small in revenue compared to most others there, but then so is STM, Broadcom and NXP in CPU revenue, so as a company it might belong there, but if counting these types of units it's architectures certainly belong on the architecture chart)
EDIT: Looking up AMD revenue numbers (from their financial filings), that data is even worse. 6.3% of a $94.2 billion market is $5.9 billion. However in 2012 earnings for AMD was $5.4 billion, and when you subtract their graphics division and others from that, you end up closer to $3.5billion to $4 billion (their GPU business alone accounts for $1billion+). So either their total estimated market size is wrong, or the AMD market share number is wrong, or they're charting something entirely different to what they say they're charting.
-
Bottomline, PPC is on it's way out.
PPC's been "on its way out" for years now, as long as you ask proponents of any other architecture. But it's taking its time about actually being "out..."
-
Like many other products it seems to die the withering under the claws of businessdepartments that don't-get-it.
Would PPC had any chance if this had been done from the start?
-
Like many other products it seems to die the withering under the claws of businessdepartments that don't-get-it.
Would PPC had any chance if this had been done from the start?
Maybe, but we'll never know. :)
When Jobs killed off the Mac clones it was the first nail in the coffin for PPC methinks.
-
" The PPC market is still larger than the x86 market in number of units shipped"
Really?
Only if you look at data from many years ago. x86 ships far more these days.
-
I wonder how many newly built washing machines still have 68k's inside them?
-
I wonder how many newly built washing machines still have 68k's inside them?
None, probably... Coldfires, yes.
-
AFAIK this isn't about PPC. PPCs have been licensable for years.
This is about POWER processors. These have never been openly licensed.
These are high end server processors, not embedded chips (though they probably are used in some very specialist IBM embedded stuff).
IBM has been pushing POWER prices down recently to compete with Intel.
Intel are going to get a lot of competition soon from ARM so Intel will want to push upwards into high end systems. This is where IBM live and they're fighting back.
-
Of course ARM dwarfs all of them, with an estimated 3 billion cpu's likely to get shipped this year...
That would be quite a disappointment given that there were nearly 2.4 billion shipped in the last quarter alone!
-
AFAIK this isn't about PPC. PPCs have been licensable for years.
This is about POWER processors. These have never been openly licensed.
These are high end server processors, not embedded chips (though they probably are used in some very specialist IBM embedded stuff).
IBM has been pushing POWER prices down recently to compete with Intel.
Intel are going to get a lot of competition soon from ARM so Intel will want to push upwards into high end systems. This is where IBM live and they're fighting back.
So the compatibility to PowerPC doesn't exist at all ?
And how is backwards compatibility to Power6,5 and 4 generation ?
The thing that gets me confused about this is that I remember that an one time Power merged with PowerPC (can't really say I understood what that means) and the thing that really makes no sense to me is how is PPC Linux compatible to these Power workstations if they are completely different things ?
-
So the compatibility to PowerPC doesn't exist at all ?
No they're compatible, POWER is generally a superset of PowerPC nowadays. This wasn't always quite the case but it has been for at least the last few iterations.
Overall I think this could be interesting, I don't think it'll drive POWER based CPUs into your desktop PC or smartphone on any real scale though. More competition across the top end for servers etc might shake things up a bit.
Everyone wondering about the design losses from the consoles needs to remember that they can just switch back again for the next generation too. Doesn't really bother or affect developers all that much, we adapt ;)
-
So the compatibility to PowerPC doesn't exist at all ?
PowerPC has big compatibility problems itself, so POWER is as pretty much compatible as they get.
You wouldn't be using the same Linux distro on a PowerPC and POWER processor anyway.
-
No they're compatible, POWER is generally a superset of PowerPC nowadays. This wasn't always quite the case but it has been for at least the last few iterations.
Hmmmm ... So G5 and PA Semis are Power4 generation (or Power5?) ...
So if somebody was to make a PowerPC CPU out of coming Power8 how
compatible would that be to Mac G5 processor ?
Are we talking difference between 68k and Coldfire or the difference between 68k and PPC or even worst ? :huh:
-
You wouldn't be using the same Linux distro on a PowerPC and POWER processor anyway.
Now I get it ...
-
Hmmmm ... So G5 and PA Semis are Power4 generation (or Power5?) ...
So if somebody was to make a PowerPC CPU out of coming Power8 how
compatible would that be to Mac G5 processor ?
Are we talking difference between 68k and Coldfire or the difference between 68k and PPC or even worst ? :huh:
The G5 is based on POWER4 with Altivec.
The PA6T is Power ISA v.2.04 (http://en.wikipedia.org/wiki/Power_Architecture#Specifications), POWER6 is Power ISA v.2.03 (http://en.wikipedia.org/wiki/Power_Architecture#Power_ISA_v.2.03).
It's more like the difference between a Cyrix 486 and a Pentium 2.
-
There more to story. I found out that IBM sold their repair part that my cousin worked for. I believe they just trying to survive.
-
The G5 is based on POWER4 with Altivec.
The PA6T is Power ISA v.2.04 (http://en.wikipedia.org/wiki/Power_Architecture#Specifications), POWER6 is Power ISA v.2.03 (http://en.wikipedia.org/wiki/Power_Architecture#Power_ISA_v.2.03).
It's more like the difference between a Cyrix 486 and a Pentium 2.
PA6T is a G5 derivative and G5 is Power4 derivative that much I do know ...
PA6T was made to be used by Apple in laptops because G5 had cooling problems but they are otherwise almost identical ...
-
Who cares if Arm outsells PowerPC and any other.
The main thing here, has to be the alliance between IBM, nVidia and Google
-
There more to story. I found out that IBM sold their repair part that my cousin worked for. I believe they just trying to survive.
Just noticed you post ...
Well no wonder they're going down after all the mistakes they made , I still can't figure out what are they doing with Power now except servers :insane:
They invented x86 and then they got cloned by far east companies and then they invented PPC and they didn't even try to make it cheaper by third party manufacturing and they lost that chance too ...
If NVidia has something to do with Power8 it means consoles (game market) , if Google wants something to do with this it means Tablets and Netbooks (Internet market) ...
Question I'm dying to get an answer to by somebody who knows a lot more of CPUs than me is will Power8 generation allow for easy migration of Morphos (and AOS 4 ) to the newly created gizmos (assuming that IBM makes them , I though they were doing Cell now) such as Power8 based game consoles , laptops and other hypothetical things ?
-
Is PPC line processors more power efficient than ARM?
-
@Blizz
IBM did not invent x86, Intel did.
Google and nvidia are interested in Power architecture for servers.
-
It may be a good thing that IBM and other giants screwed up their own business or we may have been stuck with yet another grey de facto monopoly.
-
That would be quite a disappointment given that there were nearly 2.4 billion shipped in the last quarter alone!
2.4 billion *cores* quite possibly. Because just to make this extra confusing, ARM receives royalty per core and so the number of cores often gets thrown around rather than number of CPUs. Last year they were reported to have received royalties for something like 7.9 billion cores total for the year, shipped in the region of 2.5-3 billion CPUs.
-
The main thing here, has to be the alliance between IBM, nVidia and Google
That makes it sound like they are collaborating together on a single project.
They are not.
It's more like they are members of the same club, each for their own benefit.
-
Mutal explotation :P
-
Who cares if Arm outsells PowerPC and any other.
The main thing here, has to be the alliance between IBM, nVidia and Google
Nvidia, which is an ARM licensee, is not about to become a Power licensee, says Sumit Gupta, general manager of the Tesla Accelerated Computing business unit.
http://www.theregister.co.uk/2013/08/06/ibm_opens_up_power_chips_armstyle_to_take_on_chipzilla/
-
Well that article reads so much more than just your single qoute.
Anyway, those are big players and that corporation might very well see PowerPC expanding its market share.
-
I have no doubts we'll see something interesting come out of this on a tech level, but nothing we'll actually be able to put to use ourselves. Big Iron companies are very platform agnostic - they couldn't care less if it's Intel, ARM or PPC as long as it provides massive amounts of resources within a minimal energy envelope. Using chips that use 50% less power would save them tens, if not hundreds of millions of dollars a year in operations costs.
"Google said that 100 searches are equal to a 60-watt light bulb burning for 28 minutes. " (Source: PC Mag)
All the big companies want more power with lower energy consumption. x86 provides lots of grunt, but even the most modern x86 chips consume massive amounts of power compared to ARM and such. ARM so far doesn't have adequate grunt for datacenters, but the chips are insanely efficient. There's a reason we aren't seeing ARM chips in real computers - thus far, anyways. There's also a reason why AMD and Intel are scrambling to make their offerings more power efficient.
Companies like Google wouldn't back a losing horse. It it costs them 400 million to develop a custom bred PPC chip that offers the power they need but is insanely efficient, you can bet your butt they will do it.
http://www.google.com/green/bigpicture/#/
Interesting read if anyone is interested.
-
PA6T is a G5 derivative
No, they have different heritage. It might have been designed as a G5 replacement, in the same way a 486 was a 386 replacement. Doesn't mean that one was derived from the other.
-
No, they have different heritage. It might have been designed as a G5 replacement, in the same way a 486 was a 386 replacement. Doesn't mean that one was derived from the other.
Hmmmmm , yes now I kinda see it ...
Something like Pentium and Pentium M(obile) if Pentium M was made by third party manufacturer ...
I can see Google going for the servers but I don't understand what NVidia gains by implementing Power unless they aim for the consoles ...
Is GPU somehow useful in a high end server ? :P
-
Hmmmmm , yes now I kinda see it ...
Something like Pentium and Pentium M(obile) if Pentium M was made by third party manufacturer ...
I can see Google going for the servers but I don't understand what NVidia gains by implementing Power unless they aim for the consoles ...
Is GPU somehow useful in a high end server ? :P
The whole point of project Denver was for nVidia to have a CPU+GPU combo to sell to the HPC market.
Before that, nVidia spent years trying to get a x86 license for the same purpose. And now they'll use IBM's POWER and ARM Denver....
-
The whole point of project Denver was for nVidia to have a CPU+GPU combo to sell to the HPC market.
Before that, nVidia spent years trying to get a x86 license for the same purpose. And now they'll use IBM's POWER and ARM Denver....
Ah now I understand all of it ...
Read the article again and I think it was misleading ...
So no chance in hell would be the answer to my questions :)
-
Delimiter AU: BANNED: Qld Govt outlaws new IBM contracts (http://delimiter.com.au/2013/08/07/banned-qld-govt-outlaws-new-ibm-contracts/)
IBM on the ropes? ;)
-
All the big companies want more power with lower energy consumption. x86 provides lots of grunt, but even the most modern x86 chips consume massive amounts of power compared to ARM and such. ARM so far doesn't have adequate grunt for datacenters, but the chips are insanely efficient. There's a reason we aren't seeing ARM chips in real computers - thus far, anyways. There's also a reason why AMD and Intel are scrambling to make their offerings more power efficient.
I think your not looking at the bigger ARM picture with the A15 Cortex series. The A15 could make a sweet low power consumption search engine server. https://en.wikipedia.org/wiki/ARM_Cortex-A15_MPCore
-
A15 shows promise but is still too anemic for enterprise use cache and address space wise. I have no doubts it will transform into something quite impressive in the future, though.
I've used an A15 machine in the past (Chromebook) and found it to be a hideous experience, and Google went back to Intel CPU's for the Chromebook Pixel. That being said, for the price the ARM Chromebook was well received, but I found the machine to be darned near unusable.
-
A15 shows promise but is still too anemic for enterprise use cache and address space wise. I have no doubts it will transform into something quite impressive in the future, though.
I've used an A15 machine in the past (Chromebook) and found it to be a hideous experience, and Google went back to Intel CPU's for the Chromebook Pixel. That being said, for the price the ARM Chromebook was well received, but I found the machine to be darned near unusable.
Ooh the Pixel. I want myself one of those running a proper Linux though. The screen is exquisite.
Sorry to go OT but anyone compared the screen of a Pixel with a new MBP? Which is better on the eye?
-
Nvidia presumably want to integrate POWER8 cores with their GPUs and sell compute monsters into the HPC market. Sensible licensing option for a company that is looking to expand its server market share.
I don't see Nvidia moving away from ARM for their mobile consumer SoCs though. They've invested a lot in this area - and POWER8 is just a teensy bit overkill :p
-
Nicholas:
I find the screen on the Pixel nicer than the MBP, to be honest. I give my left plum for one running a proper OS, lol. The Pixel is an absolutely gorgeous machine, but I could never get past Chrome OS.
-
A15 shows promise but is still too anemic for enterprise use cache and address space wise. I have no doubts it will transform into something quite impressive in the future, though.
The future development of the A15 (or follow on) is what I am referring to when compared to any type of future development of the POWER. I just don't see the POWER Licensing issue being taken that seriously because it's so late to the dance. Add in the rumor that significant number of POWER engineers are being shown the door in a month or two which creates a dicey for future development by IBM when ARM has a forward, but slow, development pace.
I've used an A15 machine in the past (Chromebook) and found it to be a hideous experience, and Google went back to Intel CPU's for the Chromebook Pixel. That being said, for the price the ARM Chromebook was well received, but I found the machine to be darned near unusable.
OTOH, the ARMs being used in tablets are getting pretty decent and cheap: http://hothardware.com/Reviews/Google-Nexus-7-2013-Review-Improving-Upon-the-Best-Android-Tablet/?page=1
For $229, I'm tempted in getting one.
-
Dammy,
The new Nexus 7 is an incredible little tablet. I was a pretty big fan of the iPad mini, but recently sold my Mini on Gazelle.com and got the newest Nexus 7. It's just terrific, though I wish 64 GB ones were available (or card slot would be even better).
Apple had better step up their game is all I can say. There's been very few Android contenders for the tablet throne, and the new Nexus is a better device.
-
Nicholas:
I find the screen on the Pixel nicer than the MBP, to be honest. I give my left plum for one running a proper OS, lol. The Pixel is an absolutely gorgeous machine, but I could never get past Chrome OS.
I know Linus Torvalds runs a full Linux distro on his pixel so it's certainly possible.
Whether it's possible for us mere mortals is another story though.... :)