ARM Cortex-A72 Architecture Deep Dive

[MobileEditor]

ARM's Cortex-A72 CPU adds power and performance optimizations to the previous A57 design. Here's an in-depth look at the changes to each stage in the pipeline, from better branch prediction to next-gen execution units.

ARM Cortex-A72 Architecture Deep Dive : Read more

 

[Aspiring techie]

I wonder what would happen if these guys dipped their toe into the desktop CPU market.

 

[utroz]

I wonder what would happen if these guys dipped their toe into the desktop CPU market.

Well no current Windows support so you would need to run Linux or other ARMv8 compatible OS. For people that just watch netflix, check facebook and other web pages, and type up a few papers for school or work an ARM cpu would have plenty of CPU performance..

 

[InvalidError]

I wonder what would happen if these guys dipped their toe into the desktop CPU market.

Not much.

Changing instruction set does not magically free the architecture from process limitations nor remove bottlenecks from software architecture. If ARM designed a CPU core specifically for desktop, it would hit most of the same performance scaling bottlenecks x86 has. You would likely end up with a 50W ARM chip being roughly even with a 50W Intel chip, the main difference between the two - aside from the ISA - being that the ARM chip is $100 while the Intel chip is $400.

 

[somebodyspecial]

And I think that's his point...The PRICE and how much of the public at large such a machine could get. Myself I can't wait until they put out a full desktop chip with heatsink/fan big psu, HD or SSD, 16-32GB mem etc (hopefully with an optional slot for discrete gpu when desired). As games amp up on ARM you'd most likely only be missing WINDOWS/x86 if you use pro apps stuff. Unreal4/Unity5 etc will provide nice graphics for games on the ARM side (most engines port easily today and get even better with Vulkan coming), so only pro apps would be left off for years and some of the big ones (adobe etc) might put out full apps soon anyway. Take off $200-300 for cpu and $100 for windows and I'm guessing an ARM desktop could do quite a bit of damage to WINTEL.

We see people already opting for chromebooks, tablets etc as PC's. Knock a chunk off desktop prices and you'll gains some users and push devs past mobile on arm. I'm hoping NV builds such a box at some point (just a much bigger Shield TV box really), but with multiple OS's (steamos, linux, and android) or at least a way to do it yourself. That would be a pretty versatile box It isn't so much about if they BEAT intel, as it is about dropping the price of PC's everywhere. If ARM's side wants to grow much more they have to go to desktops. Surely everyone wants a chunk of Intel's ~13B a year.

 

[viewtyjoe]

If ARM's side wants to grow much more they have to go to desktops. Surely everyone wants a chunk of Intel's ~13B a year.

ARM makes their money by licensing out their designs to other companies which manufacture the actual chips. The only company I'm aware of with the resources and ARM license to theoretically make something like this happen is AMD, and their use of ARM is more directed towards the server sector.

 

[TechyInAZ]

I wonder what would happen if these guys dipped their toe into the desktop CPU market.

I doubt that would happen, having ARM AND X86 on the desktop patform is just going to cause frustration.

 

[pug_s]

If ARM's side wants to grow much more they have to go to desktops. Surely everyone wants a chunk of Intel's ~13B a year.

ARM makes their money by licensing out their designs to other companies which manufacture the actual chips. The only company I'm aware of with the resources and ARM license to theoretically make something like this happen is AMD, and their use of ARM is more directed towards the server sector.

And I think that's his point...The PRICE and how much of the public at large such a machine could get. Myself I can't wait until they put out a full desktop chip with heatsink/fan big psu, HD or SSD, 16-32GB mem etc (hopefully with an optional slot for discrete gpu when desired). As games amp up on ARM you'd most likely only be missing WINDOWS/x86 if you use pro apps stuff. Unreal4/Unity5 etc will provide nice graphics for games on the ARM side (most engines port easily today and get even better with Vulkan coming), so only pro apps would be left off for years and some of the big ones (adobe etc) might put out full apps soon anyway. Take off $200-300 for cpu and $100 for windows and I'm guessing an ARM desktop could do quite a bit of damage to WINTEL.

We see people already opting for chromebooks, tablets etc as PC's. Knock a chunk off desktop prices and you'll gains some users and push devs past mobile on arm. I'm hoping NV builds such a box at some point (just a much bigger Shield TV box really), but with multiple OS's (steamos, linux, and android) or at least a way to do it yourself. That would be a pretty versatile box It isn't so much about if they BEAT intel, as it is about dropping the price of PC's everywhere. If ARM's side wants to grow much more they have to go to desktops. Surely everyone wants a chunk of Intel's ~13B a year.

And I think that's his point...The PRICE and how much of the public at large such a machine could get. Myself I can't wait until they put out a full desktop chip with heatsink/fan big psu, HD or SSD, 16-32GB mem etc (hopefully with an optional slot for discrete gpu when desired). As games amp up on ARM you'd most likely only be missing WINDOWS/x86 if you use pro apps stuff. Unreal4/Unity5 etc will provide nice graphics for games on the ARM side (most engines port easily today and get even better with Vulkan coming), so only pro apps would be left off for years and some of the big ones (adobe etc) might put out full apps soon anyway. Take off $200-300 for cpu and $100 for windows and I'm guessing an ARM desktop could do quite a bit of damage to WINTEL.

We see people already opting for chromebooks, tablets etc as PC's. Knock a chunk off desktop prices and you'll gains some users and push devs past mobile on arm. I'm hoping NV builds such a box at some point (just a much bigger Shield TV box really), but with multiple OS's (steamos, linux, and android) or at least a way to do it yourself. That would be a pretty versatile box It isn't so much about if they BEAT intel, as it is about dropping the price of PC's everywhere. If ARM's side wants to grow much more they have to go to desktops. Surely everyone wants a chunk of Intel's ~13B a year.

It is possible to make a hardware ARM chip that rivals Intel. The only problem is software as there is no software that would take advantage of it. Maybe in the distant future that Android OS and its apps would morph to a desktop OS that would rival to Microsoft, but not yet. Microsoft's closest adaptation of ARM soc's are windows 10 developer edition running in the Raspberry Pi 2. All of the software and games have to be written to ARM compatible. Even so, Intel's main focus now are now on low powered and mobile chips that is competing with ARM itself. Who knows, maybe by then AMD would get its act together and compete with Intel on the low powered space when they have access to 14-16nm technologies.

 

[MichaelWest]

With regard to the discussion of when someone is going to provide ARM type cpu's for desktops for ARM gaming etc we already pretty much have the beginning of this with all the many Media streaming TV boxes on the market. They mainly run Android which is not ideal for desktop applications yet and this could take many years to see this improve. Hardware wise they are more than fast enough for all the current ARM android games and there is room for them to get even faster. They don't face the same power usage limitations mobile devices face. They may be targeted for use on TV's with remotes but they can work just as well with HDMI monitors, keyboards, mice and game controllers. If all you wanted was a simple cheap desktop for email/internet and ARM gaming then they already fit the bill.

 

[cbxbiker61]

It is possible to make a hardware ARM chip that rivals Intel. The only problem is software as there is no software that would take advantage of it. Maybe in the distant future that Android OS and its apps would morph to a desktop OS that would rival to Microsoft, but not yet. Microsoft's closest adaptation of ARM soc's are windows 10 developer edition running in the Raspberry Pi 2. All of the software and games have to be written to ARM compatible. Even so, Intel's main focus now are now on low powered and mobile chips that is competing with ARM itself. Who knows, maybe by then AMD would get its act together and compete with Intel on the low powered space when they have access to 14-16nm technologies.

That is only true when you define "software" being "Windows binaries".

More and more people every day are waking up to the fact that "software" is really source code, which can be compiled on any architecture for which there is a compiler. You just have to use an open platform with an open compiler, i.e. Linux/BSD.

 

[bit_user]

Changing instruction set does not magically free the architecture from process limitations nor remove bottlenecks from software architecture. If ARM designed a CPU core specifically for desktop, it would hit most of the same performance scaling bottlenecks x86 has. You would likely end up with a 50W ARM chip being roughly even with a 50W Intel chip

If that were true, Intel could've made an Atom that's at least as efficient as competing ARM cores. But ISA does actually count for something. x86 is significantly harder to decode, and occupies more space in ICaches.

It would be interesting if someone designed an ARM v8 core for optimal single-thread performance. I'm pretty sure it could provide superior performance at the same power, and use less power at the same performance as Skylake (assuming similar design resources & process node as Intel). But this is a tall order, and there's not yet a big enough market. Maybe in 5 years, once ARM has grabbed a significant chunk of server market share, there'll be enough interest in building workstation-oriented ARM cores.

 

[bit_user]

... Now that mobile devices using the A72 are imminent ...

Imminent? Huawei Mate 8 launched in back in November, using the A72-based HiSilicon Kirin 950.

It'll be really interesting to see how Qualcomm's Kryo and NVidia's next Tegra (Parker) compare. Especially if Kryo has been outmatched before any Snapdragon 820-based phones have seen the light of day.

 

[bit_user]

That is only true when you define "software" being "Windows binaries".

More and more people every day are waking up to the fact that "software" is really source code, which can be compiled on any architecture for which there is a compiler. You just have to use an open platform with an open compiler, i.e. Linux/BSD.

Recompile from sources? That's so 1980's, dude.

Android and even .NET both support CPU-independent binaries that utilize just-in-time compilation. So, Google can ship Chromebooks based on x86, ARM, MIPS, Power, or whatever, and nearly all apps will work just fine. Only a few games & such that use the NDK might have problems.

In theory, Microsoft could do similar (and Windows 10 does have some level of ARM support), but Windows has more native apps, as you imply.

 

[cbxbiker61]

Recompile from sources? That's so 1980's, dude.

Android and even .NET both support CPU-independent binaries that utilize just-in-time compilation. So, Google can ship Chromebooks based on x86, ARM, MIPS, Power, or whatever, and nearly all apps will work just fine. Only a few games & such that use the NDK might have problems.

In theory, Microsoft could do similar (and Windows 10 does have some level of ARM support), but Windows has more native apps, as you imply.

Actually I didn't say "you" would have to compile from source. It's simply that the source code already exists, therefore a maintainer compiles for the new architecture.

Android wouldn't exist today if it wasn't able to start from a working "open source" code base.

Open source is 2000+, Windows is 1990's.

 

[bit_user]

Android wouldn't exist today if it wasn't able to start from a working "open source" code base.

Open source is 2000+, Windows is 1990's.

In terms of mainstream, maybe, but there was quite a bit of code sharing, dating back to the birth of computers. It wasn't really until the 80's that most academics and hobbyists even started copyrighting their code.

I'm all for open source, though. I'm a bit surprised there's no mainstream CPU that's opensource, by now. But I guess that's probably due to the economics of semiconductor fabrication.

 

[jonmasters]

To respond to a few misconceptions in other comments:

Open Source is an attempt to replace the term "Free Software" with something more business friendly. It dates back to the 1980s, not the 90s. And most software is still compiled code, except on modern mobile platforms, where it is heavily binary compiled code (Apple) or JVM (but not Java, Android). There are open source CPUs, but they're not going to hit prime time any time soon because the costs of building an actual high performance design without treading on patents, and also guaranteeing that the expense of silicon is going to work right within a couple of spins, coupled with the incredible cost of building a new ISA and doing all the work just to be roughly where you started, doesn't justify the effort.

 

[jimmysmitty]

Changing instruction set does not magically free the architecture from process limitations nor remove bottlenecks from software architecture. If ARM designed a CPU core specifically for desktop, it would hit most of the same performance scaling bottlenecks x86 has. You would likely end up with a 50W ARM chip being roughly even with a 50W Intel chip

If that were true, Intel could've made an Atom that's at least as efficient as competing ARM cores. But ISA does actually count for something. x86 is significantly harder to decode, and occupies more space in ICaches.

It would be interesting if someone designed an ARM v8 core for optimal single-thread performance. I'm pretty sure it could provide superior performance at the same power, and use less power at the same performance as Skylake (assuming similar design resources & process node as Intel). But this is a tall order, and there's not yet a big enough market. Maybe in 5 years, once ARM has grabbed a significant chunk of server market share, there'll be enough interest in building workstation-oriented ARM cores.

The only reason x86 is harder to decode is because of all the features it has that ARM lacks. If Intel stripped most of the features from x86 that make it so much more powerful than ARM they could easily hit the power numbers. They have gotten pretty damn close in some cases with CPUs that are still more powerful.

Either way, ARM would have a very hard time taking the desktop market due to software support. Unless they can convince the software companies to give out free copies of the software for ARM systems to customers, which they can't, they will be where most others OS and ISA stand, unable to convert the masses who hold onto everything.

 

[bit_user]

The only reason x86 is harder to decode is because of all the features it has that ARM lacks.

x86 has over 35 years worth of cruft. Each time they wanted to add instructions, they got longer and longer, because the short opcode space was already consumed. And there are multiple constants, offsets, different addressing modes, and other complexities floating around in the stew, as well.

It's pretty hard to argue that a modern ISA with all the worthwhile features of x86 would look anything like that. I think ARMv8 probably isn't too far off, feature-wise, but perhaps look at Power, if you prefer.

Probably the best acknowledgement of x86's deficiencies is IA64 - the mere fact that Intel risked a departure from the cash cow that was x86. Itanium failed for a number of reasons, but not because x86 is actually good.

And I still think it's incredibly telling that Intel hasn't resoundingly beaten ARM, given their process lead and the massive resources they've been throwing at the problem. I wonder if they'll either risk developing another proprietary ISA or possibly even designing an ARM core of their own.

 

[jimmysmitty]

The only reason x86 is harder to decode is because of all the features it has that ARM lacks.

x86 has over 35 years worth of cruft. Each time they wanted to add instructions, they got longer and longer, because the short opcode space was already consumed. And there are multiple constants, offsets, different addressing modes, and other complexities floating around in the stew, as well.

It's pretty hard to argue that a modern ISA with all the worthwhile features of x86 would look anything like that. I think ARMv8 probably isn't too far off, feature-wise, but perhaps look at Power, if you prefer.

Probably the best acknowledgement of x86's deficiencies is IA64 - the mere fact that Intel risked a departure from the cash cow that was x86. Itanium failed for a number of reasons, but not because x86 is actually good.

And I still think it's incredibly telling that Intel hasn't resoundingly beaten ARM, given their process lead and the massive resources they've been throwing at the problem. I wonder if they'll either risk developing another proprietary ISA or possibly even designing an ARM core of their own.

IA64 failed because it had to emulate x86 code and transitioning from x86 to IA64, no one was going to take that drop in performance when at the same time they could have the same x86 performance and 64bit with x86-64.

And I am not dissing ARM, but ARM itself is a low power uArch. The more features you add the more power it takes. The way ARM stays so low power is by not having a lot of the features. When ARM first came out for use in cell phones it did not have OoOE. If it did have OoOE it would have put it above the TDP spec for phone use. After time and more advanced processes cane out they were able to add in OoOE into ARM CPUs so they get better performance.

I just personally think x86 is given too much crap and people always compare it to very different applications of CPU uArchs. ARM is specifically low power. I could even compare it to the SPARC T4 which is HPC specific and has a 240W TDP but outperforms x86 in HPC applications.

As for why Intel has not beaten them, I think it is because Intel is trying to deliver a desktop experience on a phone. Atom, while much smaller than normal x86, still has most features of a desktop CPU. As well, ARM has a very strong hold and as history shows, once a technology has a hold it tends to stay for a very long time hence why we all us x86.

 

[ivyanev]

Why would anyone want an arm based Windows PC: Both AMD and Intel have some relatively low power and relatively low price CPU(APU) - Are they fast enough for browsing and watching youtube ? Sure. And ARM chip would undoubtedly be too. But how will this change the user experience at all? By using 10 watt less? I doubt it.

 

[bit_user]

Why would anyone want an arm based Windows PC

I have a ARM-based microserver (currently Raspberry Pi, but looking to upgrade). The low power consumption & passive cooling means I can leave it running 24/7 without concerns about heat, noise, dust buildup, or electricity costs. I use it mainly for media streaming, but also various automated tasks.

And I know people that use ARM-based devices as a HTPC.

That doesn't exactly answer your question, but I think some of the interest in competitive ARM-based desktop offerings is to put more pressure on Intel. If nothing else, their CPUs could be cheaper. Another benefit would be to give people more options. Since there would likely be a range of different manufacturers, perhaps there'd be some useful variation. Maybe it would push Intel to adopt new technologies faster, like PCIe v4.