AMD CPU speculation... and expert conjecture

[juanrga]

I like the new core branding:

K5 --> K6 --> K7 --> K8 --> K10 --> Bulldozer --> Piledriver --> Steamroller --> Excavator --> K12

The "K" is from Kryptonite ;-)

 

[colinp]

Another thing to note is that Carrizo is nowhere to be seen on AMD's updated roadmap. It goes from Kaveri, Beema, Mullins and Seattle in 2014 to Project Skybridge in 2015. No Excavator, no Carrizo anywhere.

With limited resources, it wouldn't surprise me if AMD has thrown all it's chips on the table (no pun intended) for Skybridge, canned Excavator, and at most we'll see a Kaveri refresh at some point to plug any gaps until Skybridge arrives, but not a full blown Excavator/Carrizo.

Are you sure?

http://benchmarkhardware.com/images/stories/2014/Mayo/semana2/ADM-Roadmap-SkyBridge-BH.JPG

Well, I took it straight from the horse's mouth from AMD's press release page and if you follow the link at the bottom of the page where it says "The core roadmap and presentation can be viewed at ir.amd.com," you get the official roadmap for graphics and cpu/apu in pdf form.

 

[szatkus]

I like the new core branding:

K5 --> K6 --> K7 --> K8 --> K10 --> Bulldozer --> Piledriver --> Steamroller --> Excavator --> K12

The "K" is from Kryptonite ;-)

More like:

K5 --> K6 --> K7 --> K8 --> K10 --> Bulldozer family --> K12

Maybe Bulldozer was K11.

 

[szatkus]

Another thing to note is that Carrizo is nowhere to be seen on AMD's updated roadmap. It goes from Kaveri, Beema, Mullins and Seattle in 2014 to Project Skybridge in 2015. No Excavator, no Carrizo anywhere.

With limited resources, it wouldn't surprise me if AMD has thrown all it's chips on the table (no pun intended) for Skybridge, canned Excavator, and at most we'll see a Kaveri refresh at some point to plug any gaps until Skybridge arrives, but not a full blown Excavator/Carrizo.

Are you sure?

http://benchmarkhardware.com/images/stories/2014/Mayo/semana2/ADM-Roadmap-SkyBridge-BH.JPG

Well, I took it straight from the horse's mouth from AMD's press release page and if you follow the link at the bottom of the page where it says "The core roadmap and presentation can be viewed at ir.amd.com," you get the official roadmap for graphics and cpu/apu in pdf form.

These roadmaps usually aren't complete. Don't kill Carrizo too early

 

[gamerk316]

Keller was very complimentary about the ARMv8 ISA having more registers. This is related to what I said about modern compiler optimizations months ago.

Keller has mentioned that the ARM core will have a "bigger engine" that the x86 core. I assume that the ARM core will be at least 6-wide (AMD FX are 4-wide).

So Itanium is obviously the way to go? Because those are two of the founding principles of the Itanic:

Registers:
128 64-bit general purpose registers
128 82-bit floating-point registers
64 1-bit predicate registers

And for execution resources?

Six general-purpose ALUs, two integer units, one shift unit
Four data cache units
Six multimedia units, two parallel shift units, one parallel multiply, one population count
Two 82-bit floating-point multiply–accumulate units, two SIMD floating-point multiply–accumulate units
Three branch units

Therefore, I conclude the Itanic will be faster then what AMD puts out, since it has moar execution units and moar registers. [/sarcasm]

Point being, lets see how nicely ARM scales up, because I suspect it will be about as ugly as X86 trying to scale down.

 

[blackkstar]

Traditional PC market is OEM pre-builts like Dell, Gateway, etc. It's very obvious that's dying. But in all honesty, that's the greatest thing for AMD to happen in the desktop market. That market is being replaced by do it yourself builders. It's easier than ever to build your own computer and there's far better online communities for it too.

Which means that, in the K8 days, remember what happened to AMD? They had a superior product but they couldn't gain the market share they deserved because Intel was bribing OEMs to not use AMD.

But guess what? It's a lot harder to bribe DIY builders. You can basically cheat in reviews and pay off reviewers but that's it. You still have people actively fighting for AMD in forums and against that sort of behaviour. Which you wouldn't have if we were still depending on traditional PCs. So really, the market shifting from traditional PC to DIY market is the best thing to happen for AMD desktop.

Also, 8350rocks and Juanrga, I think you guys are confused on what "ambidextrous" means. It means they're going to use multiple products to solve customer problems.

I'm assuming they don't want to abandon the cat cores. They are the most competitive parts with Intel. In fact, they are so good that if you want overall performance (CPU and GPU) in a low power portable, you basically want AMD by a landslide.

So, look at what AMD is proposing. ARM for server and HPC. Cat cores for consumer, and a new x86 core. If this new core is going to exist, it's not going to exist in a position to replace the functionality of an already existing product. Which is why it should be a high performance x86 core.

So AMD's product line-up would be something like

ARM for custom applications where you don't depend on Windows, HPC, LAMP servers, etc.
Cat Cores for consumer applications where power consumption is important but Windows and x86 compatibility is still needed
New x86 core for high performance x86 situations where x86 is still needed, like Windows, OSX workstations, etc.

And meanwhile, AMD will have HSA and GCN in the mix offering semi-custom solutions. So clients can come to AMD and ask for ARM + GCN + HSA and get their own custom chip. Or if Microsoft wants a new Surface chip they can go to AMD and ask for 4 Cat Cores with 2 GCN clusters and 6w TDP.

Or if another company wants to make a solid ARM tablet, they can come up and order ARM cores with GCN and HSA and have a high performance android tablet.

Look at how AMD has executed with Bulldozer. They basically wanted to create a perfect server chip and they put all their eggs in one basket. Not only was their primary core design designed for a specific workload, but it was basically their only one.

And then that chip did horrible in servers.

AMD has learned. Their ambidextrous strategy is them trying to avoid anything like that happening again, where they have one product that fails and they have to limp along on it for years.

By going cat core + arm + new x86 core, it means that if one of those fails, they still (should) have two viable products left. Which would leave them in a much better position.

Imagine if all this time AMD had Bulldozer and yet another x86 high performance core. They could have just bailed on Bulldozer much sooner and had something better. But that was never an option because they put everything in one chip.

The ARM, Cat Cores, and new x86 cores will co-exist in peace.

Juan, I have been saying all this time that AMD referring to the Piledriver Opterons as "legacy" was referring to the fact that they were old, legacy cores and the replacement wasn't here yet. There is going to be a new x86 core and it's more than likely going to fill in the performance segment that PD Opteron is leaving void.

Another thing to note is that Carrizo is nowhere to be seen on AMD's updated roadmap. It goes from Kaveri, Beema, Mullins and Seattle in 2014 to Project Skybridge in 2015. No Excavator, no Carrizo anywhere.

With limited resources, it wouldn't surprise me if AMD has thrown all it's chips on the table (no pun intended) for Skybridge, canned Excavator, and at most we'll see a Kaveri refresh at some point to plug any gaps until Skybridge arrives, but not a full blown Excavator/Carrizo.

HEDT will only happen as a happy accident - a by product of Skybridge - not by design.

Thank goodness. Someone else who gets it too. HEDT is not a goal. Designing a platform that can service HEDT and other markets is.

 

[Cazalan]

These roadmaps usually aren't complete. Don't kill Carrizo too early

He's got a point. They may have accelerated the ambidextrous plan. Then again Carrizo should have already taped out by now. They did already commit some BDVER4 code to linux.

 

[colinp]

I'm betting that if we see Carrizo, it'll be to Kaveri what Richland was to Trinity.

 

[szatkus]

I'm betting that if we see Carrizo, it'll be to Kaveri what Richland was to Trinity.

Richland and Trinity both comprised Piledriver. Frontend was bit different.
Even now we know that Carrizo's core isn't Steamroller.

 

[Cazalan]

So Itanium is obviously the way to go? Because those are two of the founding principles of the Itanic:

I'm sure some Itanic guys will love that assessment. Wonder where they would be now if HP and Oracle didn't clash heads over it and let it die.

 

[gamerk316]

So Itanium is obviously the way to go? Because those are two of the founding principles of the Itanic:

I'm sure some Itanic guys will love that assessment. Wonder where they would be now if HP and Oracle didn't clash heads over it and let it die.

The biggest issue in my mind was the compiler was sub-optimal. And the X86 performance hit certainly didn't help.

 

[8350rocks]

Juan, ARM will win in small devices...this is clear. Which is what was said to you many moons ago when you started all this mess.

x86 can do FAR too many things that would make life far more complicated for ARM to even bother with.

You are repeating again the same nonsense RISC vs CISC*, and you aren't even reading what AMD's Papermaster and Keller are saying in the conference.

Please ignore this part also:

In the news posted by our very own Usman Pirzada, we detailed how AMD will be ditching their existing modular architecture and go for the Simultaneous Multi Threading (SMT) design.

[...]

With Jim Keller onboard, AMD is finally developing a new core architecture that’s built from scratch and means that AMD won’t make the same mistakes they made with Bulldozer. The architecture is still far from being ready and we are looking at a launch scenario of atleast 2016-2017 so that’s still a long time to go from where we are currently.

[...]

AMD’s 2015 line of APUs will stick with Excavator which will be the last modular AMD core before it hops onboard the SMT design.

* I already explained you that the first "C" in CISC doesn't mean "Complex", that any modern x86 CPU is implemented over RISC-like uops and that ARMv8 is a better ISA than x86-64.

*SIGH* this has nothing to do with anything you posted. x86 as an ISA is one thing, ARM is another. Can they accomplish the same tasks? Mostly, just in different ways...ARM would likely have to implement a few additional libraries/indexes to encompass everything, but that is not to say they could not.

The difference is in intent for design. ARM is intended to be ULP primarily, and is the chief advantage of the uarch. x86 was designed to be a powerful all encompassing uarch for EVERY level of compute. In order for ARM to do what x86 does, you end up drawing more power to do the same tasks. x86 has been at the high performance game FAR longer than ARM, and ARM would have to go through all the growing pains of making the uarch engineered to do what x86 does and bring about the refinements.

If ARM wanted to compete with x86 in HEDT, or anything outside ULP, then it would literally lose 20 years of progress. ARM is not a uarch designed for such tasks...and would end up losing more ground than they could ever hope to gain.

You do not understand, the lessons x86 learned along the way took a LONG time. ARM cannot do it over night...even knowing what they went through with x86 designs.

Please post something relevant.

 

[8350rocks]

Traditional PC market is OEM pre-builts like Dell, Gateway, etc. It's very obvious that's dying. But in all honesty, that's the greatest thing for AMD to happen in the desktop market. That market is being replaced by do it yourself builders. It's easier than ever to build your own computer and there's far better online communities for it too.

Which means that, in the K8 days, remember what happened to AMD? They had a superior product but they couldn't gain the market share they deserved because Intel was bribing OEMs to not use AMD.

But guess what? It's a lot harder to bribe DIY builders. You can basically cheat in reviews and pay off reviewers but that's it. You still have people actively fighting for AMD in forums and against that sort of behaviour. Which you wouldn't have if we were still depending on traditional PCs. So really, the market shifting from traditional PC to DIY market is the best thing to happen for AMD desktop.

Also, 8350rocks and Juanrga, I think you guys are confused on what "ambidextrous" means. It means they're going to use multiple products to solve customer problems.

I'm assuming they don't want to abandon the cat cores. They are the most competitive parts with Intel. In fact, they are so good that if you want overall performance (CPU and GPU) in a low power portable, you basically want AMD by a landslide.

So, look at what AMD is proposing. ARM for server and HPC. Cat cores for consumer, and a new x86 core. If this new core is going to exist, it's not going to exist in a position to replace the functionality of an already existing product. Which is why it should be a high performance x86 core.

So AMD's product line-up would be something like

ARM for custom applications where you don't depend on Windows, HPC, LAMP servers, etc.
Cat Cores for consumer applications where power consumption is important but Windows and x86 compatibility is still needed
New x86 core for high performance x86 situations where x86 is still needed, like Windows, OSX workstations, etc.

And meanwhile, AMD will have HSA and GCN in the mix offering semi-custom solutions. So clients can come to AMD and ask for ARM + GCN + HSA and get their own custom chip. Or if Microsoft wants a new Surface chip they can go to AMD and ask for 4 Cat Cores with 2 GCN clusters and 6w TDP.

Or if another company wants to make a solid ARM tablet, they can come up and order ARM cores with GCN and HSA and have a high performance android tablet.

Look at how AMD has executed with Bulldozer. They basically wanted to create a perfect server chip and they put all their eggs in one basket. Not only was their primary core design designed for a specific workload, but it was basically their only one.

And then that chip did horrible in servers.

AMD has learned. Their ambidextrous strategy is them trying to avoid anything like that happening again, where they have one product that fails and they have to limp along on it for years.

By going cat core + arm + new x86 core, it means that if one of those fails, they still (should) have two viable products left. Which would leave them in a much better position.

Imagine if all this time AMD had Bulldozer and yet another x86 high performance core. They could have just bailed on Bulldozer much sooner and had something better. But that was never an option because they put everything in one chip.

The ARM, Cat Cores, and new x86 cores will co-exist in peace.

Juan, I have been saying all this time that AMD referring to the Piledriver Opterons as "legacy" was referring to the fact that they were old, legacy cores and the replacement wasn't here yet. There is going to be a new x86 core and it's more than likely going to fill in the performance segment that PD Opteron is leaving void.

Another thing to note is that Carrizo is nowhere to be seen on AMD's updated roadmap. It goes from Kaveri, Beema, Mullins and Seattle in 2014 to Project Skybridge in 2015. No Excavator, no Carrizo anywhere.

With limited resources, it wouldn't surprise me if AMD has thrown all it's chips on the table (no pun intended) for Skybridge, canned Excavator, and at most we'll see a Kaveri refresh at some point to plug any gaps until Skybridge arrives, but not a full blown Excavator/Carrizo.

HEDT will only happen as a happy accident - a by product of Skybridge - not by design.

Thank goodness. Someone else who gets it too. HEDT is not a goal. Designing a platform that can service HEDT and other markets is.

I mostly agree with this, except I see the ARM solutions being aimed specifically at ULP and "microservers" as loose a definition as that tends to have...

 

[8350rocks]

Traditional PC market is OEM pre-builts like Dell, Gateway, etc. It's very obvious that's dying. But in all honesty, that's the greatest thing for AMD to happen in the desktop market. That market is being replaced by do it yourself builders. It's easier than ever to build your own computer and there's far better online communities for it too.

Which means that, in the K8 days, remember what happened to AMD? They had a superior product but they couldn't gain the market share they deserved because Intel was bribing OEMs to not use AMD.

But guess what? It's a lot harder to bribe DIY builders. You can basically cheat in reviews and pay off reviewers but that's it. You still have people actively fighting for AMD in forums and against that sort of behaviour. Which you wouldn't have if we were still depending on traditional PCs. So really, the market shifting from traditional PC to DIY market is the best thing to happen for AMD desktop.

Also, 8350rocks and Juanrga, I think you guys are confused on what "ambidextrous" means. It means they're going to use multiple products to solve customer problems.

I'm assuming they don't want to abandon the cat cores. They are the most competitive parts with Intel. In fact, they are so good that if you want overall performance (CPU and GPU) in a low power portable, you basically want AMD by a landslide.

So, look at what AMD is proposing. ARM for server and HPC. Cat cores for consumer, and a new x86 core. If this new core is going to exist, it's not going to exist in a position to replace the functionality of an already existing product. Which is why it should be a high performance x86 core.

So AMD's product line-up would be something like

ARM for custom applications where you don't depend on Windows, HPC, LAMP servers, etc.
Cat Cores for consumer applications where power consumption is important but Windows and x86 compatibility is still needed
New x86 core for high performance x86 situations where x86 is still needed, like Windows, OSX workstations, etc.

And meanwhile, AMD will have HSA and GCN in the mix offering semi-custom solutions. So clients can come to AMD and ask for ARM + GCN + HSA and get their own custom chip. Or if Microsoft wants a new Surface chip they can go to AMD and ask for 4 Cat Cores with 2 GCN clusters and 6w TDP.

Or if another company wants to make a solid ARM tablet, they can come up and order ARM cores with GCN and HSA and have a high performance android tablet.

Look at how AMD has executed with Bulldozer. They basically wanted to create a perfect server chip and they put all their eggs in one basket. Not only was their primary core design designed for a specific workload, but it was basically their only one.

And then that chip did horrible in servers.

AMD has learned. Their ambidextrous strategy is them trying to avoid anything like that happening again, where they have one product that fails and they have to limp along on it for years.

By going cat core + arm + new x86 core, it means that if one of those fails, they still (should) have two viable products left. Which would leave them in a much better position.

Imagine if all this time AMD had Bulldozer and yet another x86 high performance core. They could have just bailed on Bulldozer much sooner and had something better. But that was never an option because they put everything in one chip.

The ARM, Cat Cores, and new x86 cores will co-exist in peace.

Juan, I have been saying all this time that AMD referring to the Piledriver Opterons as "legacy" was referring to the fact that they were old, legacy cores and the replacement wasn't here yet. There is going to be a new x86 core and it's more than likely going to fill in the performance segment that PD Opteron is leaving void.

Another thing to note is that Carrizo is nowhere to be seen on AMD's updated roadmap. It goes from Kaveri, Beema, Mullins and Seattle in 2014 to Project Skybridge in 2015. No Excavator, no Carrizo anywhere.

With limited resources, it wouldn't surprise me if AMD has thrown all it's chips on the table (no pun intended) for Skybridge, canned Excavator, and at most we'll see a Kaveri refresh at some point to plug any gaps until Skybridge arrives, but not a full blown Excavator/Carrizo.

HEDT will only happen as a happy accident - a by product of Skybridge - not by design.

Thank goodness. Someone else who gets it too. HEDT is not a goal. Designing a platform that can service HEDT and other markets is.

I mostly agree with this, except I see the ARM solutions being aimed specifically at ULP and "microservers" as loose a definition as that tends to have...

 

[harly2]

Isn't it

I like the new core branding:

K5 --> K6 --> K7 --> K8 --> K10 --> Bulldozer --> Piledriver --> Steamroller --> Excavator -->What ever new x86 is

A57-- > K12

The "K" is from Kryptonite ;-)

 

[juanrga]

I like the new core branding:

K5 --> K6 --> K7 --> K8 --> K10 --> Bulldozer --> Piledriver --> Steamroller --> Excavator --> K12

The "K" is from Kryptonite ;-)

More like:

K5 --> K6 --> K7 --> K8 --> K10 --> Bulldozer family --> K12

Maybe Bulldozer was K11.

Yes Bulldozer is sometimes referred as K11.

 

[juanrga]

Keller was very complimentary about the ARMv8 ISA having more registers. This is related to what I said about modern compiler optimizations months ago.

Keller has mentioned that the ARM core will have a "bigger engine" that the x86 core. I assume that the ARM core will be at least 6-wide (AMD FX are 4-wide).

So Itanium is obviously the way to go? Because those are two of the founding principles of the Itanic:

Registers:
128 64-bit general purpose registers
128 82-bit floating-point registers
64 1-bit predicate registers

And for execution resources?

Six general-purpose ALUs, two integer units, one shift unit
Four data cache units
Six multimedia units, two parallel shift units, one parallel multiply, one population count
Two 82-bit floating-point multiply–accumulate units, two SIMD floating-point multiply–accumulate units
Three branch units

Therefore, I conclude the Itanic will be faster then what AMD puts out, since it has moar execution units and moar registers. [/sarcasm]

Point being, lets see how nicely ARM scales up, because I suspect it will be about as ugly as X86 trying to scale down.

Let us check the Cyclone core. It is wide, about 50% wider than cores in AMD FX and has more registers. And how does it perform? Well two cyclone cores @1.3GHz are enough to match quad-cores from Intel/AMD @ 1.5GHz and higher frequencies.

I knew you are an expert in 'mantel', but I didn't know that you are also expert in CPU architecture to discuss sarcastically Keller tech. decisions about its next core. You have impressed me again!

 

[juanrga]

So, look at what AMD is proposing. ARM for server and HPC. Cat cores for consumer, and a new x86 core.

No.

On the slide above we see that AMD is not only targeting servers and embedded markets, but also ultra low power client devices for its 64-bit ARM designs (presumably notebooks, chromebooks, tablets).

Juan, I have been saying all this time that AMD referring to the Piledriver Opterons as "legacy" was referring to the fact that they were old, legacy cores and the replacement wasn't here yet. There is going to be a new x86 core and it's more than likely going to fill in the performance segment that PD Opteron is leaving void.

I have been for months saying that no new FX-Steamroller/Excavator CPU was coming because AMD was preparing a replacement. I also said that Warsaw CPU was released only for legacy customers. I also said that the FX- Piledriver series was extended to 2015 because it was being replaced in 2016 by a new APU made by both Keller and Koduri. I said Keller was designing new CPU arch and Koduri a new GPU arch. We know now a bit more about Keller work. Future conferences will say that Koduri is doing.

 

[juanrga]

*SIGH* this has nothing to do with anything you posted. x86 as an ISA is one thing, ARM is another. Can they accomplish the same tasks? Mostly, just in different ways...ARM would likely have to implement a few additional libraries/indexes to encompass everything, but that is not to say they could not.

The difference is in intent for design. ARM is intended to be ULP primarily, and is the chief advantage of the uarch. x86 was designed to be a powerful all encompassing uarch for EVERY level of compute. In order for ARM to do what x86 does, you end up drawing more power to do the same tasks. x86 has been at the high performance game FAR longer than ARM, and ARM would have to go through all the growing pains of making the uarch engineered to do what x86 does and bring about the refinements.

If ARM wanted to compete with x86 in HEDT, or anything outside ULP, then it would literally lose 20 years of progress. ARM is not a uarch designed for such tasks...and would end up losing more ground than they could ever hope to gain.

You do not understand, the lessons x86 learned along the way took a LONG time. ARM cannot do it over night...even knowing what they went through with x86 designs.

You continue confused. ARM has been optimized for efficiency during years. This is why you need two x86 cores from AMD/Intel to match the performance of a single ARM core from Apple/Nvidia.

I see you continue negating the comments made by Keller, Papermaster, and anyone at AMD conference, when they said that K12 will be a "high-performance core and" and I see you continue ignoring Keller talk explaining why he can extract more performance from ARM than from x86.

 

[jdwii]

Na all things considered the new x86 processor jim is making is pretty interesting and i hope we know more soon.

 

[8350rocks]

*SIGH* this has nothing to do with anything you posted. x86 as an ISA is one thing, ARM is another. Can they accomplish the same tasks? Mostly, just in different ways...ARM would likely have to implement a few additional libraries/indexes to encompass everything, but that is not to say they could not.

The difference is in intent for design. ARM is intended to be ULP primarily, and is the chief advantage of the uarch. x86 was designed to be a powerful all encompassing uarch for EVERY level of compute. In order for ARM to do what x86 does, you end up drawing more power to do the same tasks. x86 has been at the high performance game FAR longer than ARM, and ARM would have to go through all the growing pains of making the uarch engineered to do what x86 does and bring about the refinements.

If ARM wanted to compete with x86 in HEDT, or anything outside ULP, then it would literally lose 20 years of progress. ARM is not a uarch designed for such tasks...and would end up losing more ground than they could ever hope to gain.

You do not understand, the lessons x86 learned along the way took a LONG time. ARM cannot do it over night...even knowing what they went through with x86 designs.

You continue confused. ARM has been optimized for efficiency during years. This is why you need two x86 cores from AMD/Intel to match the performance of a single ARM core from Apple/Nvidia.

I see you continue negating the comments made by Keller, Papermaster, and anyone at AMD conference, when they said that K12 will be a "high-performance core and" and I see you continue ignoring Keller talk explaining why he can extract more performance from ARM than from x86.

First, it is you who are confused, you need more than 12 ARM processors from Apple or NVidia to compete with 1 single i7-4930k. Show me ONE x86 metric where ARM even COMPETES 1 to 1.

Also, you continue trying to shift the conversation off topic.

ARM will not scale well upward. Period.

I care not what Papermaster said...you take what Keller said out of context, and Rory Read knows less about APU/CPU architecture than he knows about Tarot Card readings. He came from Lenovo, he knows how to run a business, which is his job...he is not an Engineer, nor was he ever.

Now, please stay on topic and address my points.

 

[bmacsys]

Juan, ARM will win in small devices...this is clear. Which is what was said to you many moons ago when you started all this mess.

x86 can do FAR too many things that would make life far more complicated for ARM to even bother with. Could ARM try? Sure...would they unseat x86 for this? Not very likely...perfect example is Windows RT. People will want windows on x86, why? Because it does all the things they want. Additionally...you will never get consumer Joe Q Public to switch his perfectly fine PC at home for some "new fangled" device just because a promo slide you showed by some speculator says ARM will win. He does not want to buy a new x86 rig to do what he wants better, so why would he abandon all that familiarity to go to something "untried" in the public PC sector? Are they fine with it in a tablet? Sure...would they trust their jobs and lives to it? Well, let us just say, FAR too many people are still on XP. The proof is in the pudding...ARM will not win anything that it is not already winning. I might add that x86 is already trying very flagrantly to steal market share as well.

I would argue, if x86 hardware competes at PPW for ULP devices...ARM dies a quick death. Look at the number of new tablets now that are running x86 compared to what was offered a year ago even. Then, consider many x86 devices can run android because it is linux, essentially, which means that the software would not have any issues adapting. That is all people really care about...will this play angry birds, and will it run android? If the answer is yes, then there is no concern from the public.

I think you are wrong. Little by little ARM devices are eating away at x86. PC sales continue to fall. There is real momentum being built. ARM device sales dwarf x86 device sales right now. A few years ago people would have laughed at this ever happening. I think eventually Apple will build an ARM powered MacBook. It will sell. Then you will see all the big Asian company's follow suit by building cheap Android desktops. Microsoft knows this. That is why they built the Surface RT. Intel will be a bit player in 10 years. ARM is winning the battle right now. Intel gets huge margins with x86. Those kinds of margins don't exist in the ARM world. Processors are $20 not $200. Intel can't survive selling $20 processors.

 

[8350rocks]

Juan, ARM will win in small devices...this is clear. Which is what was said to you many moons ago when you started all this mess.

x86 can do FAR too many things that would make life far more complicated for ARM to even bother with. Could ARM try? Sure...would they unseat x86 for this? Not very likely...perfect example is Windows RT. People will want windows on x86, why? Because it does all the things they want. Additionally...you will never get consumer Joe Q Public to switch his perfectly fine PC at home for some "new fangled" device just because a promo slide you showed by some speculator says ARM will win. He does not want to buy a new x86 rig to do what he wants better, so why would he abandon all that familiarity to go to something "untried" in the public PC sector? Are they fine with it in a tablet? Sure...would they trust their jobs and lives to it? Well, let us just say, FAR too many people are still on XP. The proof is in the pudding...ARM will not win anything that it is not already winning. I might add that x86 is already trying very flagrantly to steal market share as well.

I would argue, if x86 hardware competes at PPW for ULP devices...ARM dies a quick death. Look at the number of new tablets now that are running x86 compared to what was offered a year ago even. Then, consider many x86 devices can run android because it is linux, essentially, which means that the software would not have any issues adapting. That is all people really care about...will this play angry birds, and will it run android? If the answer is yes, then there is no concern from the public.

I think you are wrong. Little by little ARM devices are eating away at x86. PC sales continue to fall. There is real momentum being built. ARM device sales dwarf x86 device sales right now. A few years ago people would have laughed at this ever happening. I think eventually Apple will build an ARM powered MacBook. It will sell. Then you will see all the big Asian company's follow suit by building cheap Android desktops. Microsoft knows this. That is why they built the Surface RT. Intel will be a bit player in 10 years. ARM is winning the battle right now. Intel gets huge margins with x86. Those kinds of margins don't exist in the ARM world. Processors are $20 not $200. Intel can't survive selling $20 processors.

Because ARM cores go into disposable devices like tablets and phones.

The fact that x86 is stagnating is not because ARM is "winning", it is because the PC market is saturated. No amount of ARM sales that have been made to date are erroding any part of any segment of x86 market share, the 2 do not compete except in very small portions of niche segments (i.e. convertibles).

x86 is honestly gaining ground in those markets where ARM is traditionally dominant as well. Granted, it is not gaining market share at 2 to 1, but when you are talking about 100s of millions of devices even 2-3% gain is a substantial amount of devices that are now x86 that were not 12 months ago.

 

[palladin9479]

Point being, lets see how nicely ARM scales up, because I suspect it will be about as ugly as X86 trying to scale down.

ARM's amazing performance per watt is due to the simplicity of the design. The ISA itself is irrelevant but the design's that accompany that ISA tend to have a minimalist mindset. When you look at an x86 CPU core you realize that the part of the core that does 90%+ of the work only takes up 10~12% of the space used. The rest of the die space is used to accelerate, optimize and increase throughput to that 10~12%. ARM isn't unique, it faces the exact same design challenges that the x86 CPU's do, in order to scale up in per-core performance the would need to add more robust predictors, dispatchers, schedulers and cache logic. You start adding tons of additional units in each core as a way to enable the performance to ramp up, but those extra components take up space and require power which blows out your cost / power budget. The only way around this is massive parallelism, chaining together dozens of smaller cheaper weak cores instead of a handful of powerful ones.

 

[de5_Roy]

"First, its important to differentiate between HSA on the desktop and HSA on mobile. In the mobile world, HSA and heterogeneous compute seem to have moved much further along than in the desktop space. MediaTek (one of the HSA Foundation’s founding members) is making great progress and has some clear wins in the space with things like CorePilot, its Heterogeneous Multi-Processing (HMP) control software for big.LITTLE. In contrast, AMD has not had the same sort of success in the desktop world."
http://vr-zone.com/articles/amd-arm-hsa-skybridge/76959.html
lack of marketshare will do that to you.