AMD CPU speculation... and expert conjecture

[jdwii]

It is worth mentioning that Windows RT fiasco has little to do with ARM, but is due to Microsoft mismanagement. The same mismanagement behind PS4 outperforming XB1 or linux destroying windows server

SA has an interesting article about all that

http://semiaccurate.com/2014/05/15/microsoft-now-irrelevant-computing-want-know/

I find a lot of fallacious arguments in that article

 

[jdwii]

Juanrga....

Really enjoy lots of your points, thanks for the info you have posted. But if I may....I notice over at semi-accurate you don't do all the self promotion with all the "as I predicted" and "like I said"... I I I. Would you mind talking more like that here. Not sure if your using this forum as your own for an ego boost or something but Its getting to me. I go around every few days and read all the AMD news and just keep running into you doing this here. You quote someone and then rant about how you said it. Everything you find is available at numerous sites. And if you do predict something and its right that's cool, know one is viewing you as an oracle or something, just take or give info without the self promotion please. Again thanks for the digging and the info but its clear that you do this for yourself, don't pretend your doing the world a favor. Its your hobby and that's cool, in this forum on this site do you mind just stopping the constant self promotion. Not sure if you think some of the old semi-accurate guys would put you in your place or something but you definitely talk in a different tone here like everyone should bow down.

Completely agree with this sentiment.

To be fair, though, Juanrga isn't the only one around here with a huge ego and who keeps repeating the old "I was right, bow down to my superior intellect". Gamerk and 8350rocks have both been repeat offenders, maybe there are others. This thread would be all the nicer if everyone would cut it out.

We can also throw noob in the mix, Gamerk and Noob will never ever give the slightest credit to AMD, Juan and 8350Rocks just keep throwing predictions, the only difference between them is that Juan smear it into everyone face`s when he`s right, and he usually is... while 8350Rocks seems to fail most of his predictions due to his AMD insider constantly giving him massively wrong info.

Name certain cases i can name plenty where he was wrong and yet the mods let him continue to put others down, still waiting on his whole A10 comparable to an I5 thing in CPU benchmarks on average based on the majority, not a selective few

 

[colinp]

It is worth mentioning that Windows RT fiasco has little to do with ARM, but is due to Microsoft mismanagement. The same mismanagement behind PS4 outperforming XB1 or linux destroying windows server

SA has an interesting article about all that

http://semiaccurate.com/2014/05/15/microsoft-now-irrelevant-computing-want-know/

I find a lot of fallacious arguments in that article

Semiaccurate, natch.

 

[jed]

^^

Couldnt agree more. Stroking juans ego with " he usually is" is just going to make this worse. Hes been allowed to insult people without warning because he does it in a way thats not blatantly obvious so they let it slide.

For example I don't know how to read, never listen to him, don't know anything about computers, don't know how to use math, etc. according to Juan. He always tries to claim I said something I never said in order to prove "he is right all the time". And yes I was banned for 2 weeks because I called him out on it.

as for claiming I will never give credit to AMD read my other posts. Im an avid AMD fan. Im also a realist, AMD abandoning high end for lower performance APU & ARM is a mistake. If the fx-8350 is the fastest thing AMD releases in the next 5 years, what purchase option do I have? Im not going to downgrade to an A10 quad core to run my 290x. AMD can't compete with samsung, qualcomm, & Nvidia in the ARM market. They aren't going to let AMD just take over the market without a fight, a fight I doubt AMD can afford. The 3 I mentioned can drop 1B on marketing without putting a dent in their budget. The only thing AMD can do is what they are doing, marketing hype campaigns to push their superiority before the product is completed.

I was about to inform him that you have been one of the strongest voices for AMD in this thread,
so for him to make that statement he didn't start read this thread from the beginning.

 

[juanrga]

The chart above shows the TAM (Total Addressable Market) for processors across embedded, mobile, PC, console, and server segments by architecture. The gray bars in the middle of the chart lump together Power, Sparc, MIPS, various mainframe, and other proprietary chips, this is falling. The orange bars show the drop of the x86 market, while the purple bars, representing ARM chips, show the exponential rise of this architecture.

By about 2018 ARM chips will represent about a 50% of AMD TAM, the other 50% is x86. AMD new ambidextrous strategy is an adaptation to this market situation. If we extrapolate the chart beyond 2018 we can see ARM winning over x86; this is in the "long run", it is not happening tomorrow.

 

[-Fran-]

Yes and it was slower than anything now image that on a server controlling the login process of 500 computers.

Well, if you or your boss are thinking about using a Tablet ARM CPU/SoC to run a server farm, well... You guys need to revisit your ideas.

Don't mix the topics, since they're very different.

Windows RT and software "portability" is one topic on its own, and how the ARMv7 and ARMv8 with their consequent CPU uAarchs are able to handle themselves against Intel and AMD in the vertical markets is another topic.

In news that could be related to this thread (well, not really, haha), new laws for trade "fairness" were ruled a couple of years ago (I found about this through our mandatory training in such topics, here at the company, lol). They're more client (or end user) focused more than Company interaction, but they're a good read if you guys want to take a look at it: Dodd-Frank Law

Cheers!

 

[de5_Roy]

In news that could be related to this thread

since amd will be making multiple cpus with a57 and multiple skus of each before moving on to k12 after project skybridge, arm v8 and a57 may be related as long as the amd context is correctly applied. these include hierofalcon, seattle, whatever comes next. i wanna know if amd overhauled the bd/jag iommu for project skybridge. k12 is long way out, seattle is just a bunch of a57 cores slapped together with massive cache and low-ish clockrate (assuming from the seattle server board's heatsink). project skybridge seems interesting. the less-than-subtle jab at intel also helps...

 

[juanrga]

Yes and it was slower than anything now image that on a server controlling the login process of 500 computers.

Well, if you or your boss are thinking about using a Tablet ARM CPU/SoC to run a server farm, well... You guys need to revisit your ideas.

Don't mix the topics, since they're very different.

Windows RT and software "portability" is one topic on its own, and how the ARMv7 and ARMv8 with their consequent CPU uAarchs are able to handle themselves against Intel and AMD in the vertical markets is another topic.

The PS4 architect tells the anecdote that during the hardware development phase one prominent boss at Sony shared his disagreement by the choice of x86 architecture, because this boss claimed that x86 was an inadequate architecture for gaming. I suppose that one can find ignorant bosses everywhere.

I mentioned before that Microsoft has participated in the development of the new SBSA (Server Base System Architecture) used in AMD Seattle SoC. It seems this important point was ignored.

Those the words of Microsoft representative about SBSA:

http://www.computerworld.com/s/article/9245854/Microsoft_joins_ARM_server_effort

We can confirm Microsoft participates in an industry forum driven by ARM around the Server Base System Architecture, as a part of our continuous work with industry partners to deliver solutions our customers want

Some rumors say Microsoft has already ported Windows server to ARM and engineers are playing with it behind the doors.

 

[jimmysmitty]

http://www.computingcompendium.com/p/arm-vs-intel-benchmarks.html

Trying to say Intel's Atom = haswell is the biggest flawed statement in this whole ARM vs x86 discussion. Atom is Intel's attempt to make an efficient x86 (aka V8) by neutering it.

ARM has a long way to go to catch that 16,774 score, at best its currently less than 3000. Even single core @ 1.3 ghz each, the infamous apple A7 is less than half of what haswell is and Atom being 1/3 as fast.

But hey, its faster than Atom, lets keep pretending that Atom = high end.

I don't see anyone in this thread pretending that Atom is high end. I don't see anyone in this thread comparing K12 to Atom. I see comparisons of K12 to Piledriver and Haswell desktops.

You comparing the performance of a 1.3GHz phone chip (~2W) against a 3.5GHz desktop chip (80W) and claiming "ARM has a long way to go to catch" is so useless and wrong as comparing the performance of a 1.3GHz i3-4010Y (11.5W) against an 3.5GHz i7-4960x (130W) and claiming that "Haswell has a long way to go to catch Sandy Bridge".

You are confounding architectures and ISAs, with specific implementations at particular products.

A serious analysis of cyclone microarchitecture compared to Haswell was made by Anand

http://www.anandtech.com/show/7910/apples-cyclone-microarchitecture-detailed

An exhaustive analysis of armv8 ISA is given by Kanter. Keller has also commented which are the strong points of the ISA.

Finally, it is worth mention AMD Seattle is already a 25W chip. Evidently Seattle is not designed for phones but for microservers.

The only comparison to Haswell I see in that article is to the 192 micro-ops that both Haswell and Apples Cyclone have and the same branch mispredict. Other than that it is more of a comparison to Swift. I don't think Anand is dumb enough to come out saying that Apples first 64bit self(somewhat) designed ARM based CPU is superior to a x86 CPU.

And no one has said anything about comparing Haswell to ARM. I have already said Haswell is a much more complex chip with superior IPC and will eat it alive. That is expected as ARM has never been about the super complexity that x86 has achieved and which is why they are able to keep that nice and low TDP. That is also why Intel took a Pentium P56C core and stripped it down to very basic and then called it Atom trying to push a 2w TDP.

Once ARM starts to throw in that same level of complexity as x86 to achieve the same performance on all levels, that's when the ARM TDP advantage will go away. But as long as it stays more simplified to keep a lower TDP it will never overtake x86 performance.

From the EETIMES article I mentioned before:

Keller suggested AMD will consolidate its work on separate Bulldozer and Jaguar x86 cores into one processor line in the future.

Kind of scarey if you think about it. ARM can't do it, Intel can't do it, but somehow AMD will do it. A one size fits all core. The holy grail of computing.

There sure were some harsh comments regarding Keller on there.

pretty much sais it all right there.

Seems like AMD is banking a lot on ARM becoming more dominant.

I like the "substantial share of the x86 CPU market" bit. If Rory has his way, they wont compete with Intel the same way so they will never look to design superior x86 CPUs. How do they plan to take more x86 market share, or substantial, if they do not ever design a superior uArch?

So if ARM replaces x86... what about PC Gaming and all the x86 Games? New Tech is always cool, but what we gona do when all PCs are ARM... 98% of software is on x86.

I don`t mind to eventually move into ARM as long as ARM manages to run (not emulate) x86 software, i care about my PC games everything else is secondary for me.

Smart companies will develop games with cross platform APIs. Unreal engines 3 and 4 can both run in a web browser even.

And how many game companies will willingly spend the time and money to port the old games over? Except for VALVe, I don't see the other companies doing it. Oh I know what they will do, they will charge you $5-$20 bucks on some market and make it like a retro game. Sort of like how you can buy, again, old Xbox or PS2 games on their respective markets to play.

It is worth mentioning that Windows RT fiasco has little to do with ARM, but is due to Microsoft mismanagement. The same mismanagement behind PS4 outperforming XB1 or linux destroying windows server

SA has an interesting article about all that

http://semiaccurate.com/2014/05/15/microsoft-now-irrelevant-computing-want-know/

I like the part talking about the more secure free alternatives. I hope they realize that the more popular those free alternatives get the more unsecure they become. Look at Apple. Mac OSX has had a big jump in viruses. Android, a Linux based OS, has multiple viruses now and multiple anti-virus applications. It is the nature of the beast. If Microsoft becomes smaller it will just become more secure as the hackers and virus makers will move on to the bigger target.

Let me ask you. Is t more profitable for the Cryptolocker virus to attack Mac OSX, Linux or Windows?

Spoiler

jWindows can't be in "ARM PCs", because there are no ARM CPU vendors like AMD or Intel so consumers can actually assemble one. AMD wants that to change, and it looks very promising.

Spoiler

Not only AMD. Nvidia also want ARM from phones to supercomputers:

ARM is already the standard architecture for mobile devices. Project Denver extends the range of ARM systems upward to PCs, data center servers, and supercomputers. ARM’s modern architecture, open business model, and vibrant eco-system have led to its pervasiveness in cell phones, tablets, and other embedded devices. Denver is the catalyst that will enable these same factors to propel ARM to become pervasive in higher-end systems.

Denver frees PCs, workstations and servers from the hegemony and inefficiency of the x86 architecture. For several years, makers of high-end computing platforms have had no choice about instruction-set architecture. The only option was the x86 instruction set with variable-length instructions, a small register set, and other features that interfered with modern compiler optimizations, required a larger area for instruction decoding, and substantially reduced energy efficiency.

Denver provides a choice. System builders can now choose a high-performance processor based on a RISC instruction set with modern features such as fixed-width instructions, predication, and a large general register file. These features enable advanced compiler techniques and simplify implementation, ultimately leading to higher performance and a more energy-efficient processor.

Mainly because NVidia was not able to get a x86 license. Do you remember the debacle when Intel went from FSB to IMC? NVidia didn't have a license that covered anything beyond Intels FSB design for chipsets. So NVidia lost their ability to make chipsets. Thank God as their chipsets were just bad.

After that I am sure NVidia wanted to branch out and would have killed for a x86 license. I am sure they still would.

 

[-Fran-]

Some rumors say Microsoft has already ported Windows server to ARM and engineers are playing with it behind the doors.

Have some links to share about that? I could ask the tech guys in our server division, but if they're under NDA, I really doubt they'll tell me anything, lol.

Mainly because NVidia was not able to get a x86 license. Do you remember the debacle when Intel went from FSB to IMC? NVidia didn't have a license that covered anything beyond Intels FSB design for chipsets. So NVidia lost their ability to make chipsets. Thank God as their chipsets were just bad.

After that I am sure NVidia wanted to branch out and would have killed for a x86 license. I am sure they still would.

I agree with you there, but nVidia has been very stupid about some key things when dealing with Intel. I remember reading that Mr. Huang, when there was talk about a possible Intel buy-out/merger with nVidia, they were very arrogant. I'm still wondering if all that was true or not, but I've always been under the impression that nVidia is sitting on their own little island because they, as an organization, are very stubborn and proud, but in a bad way.

Oh well, this doesn't get more than just old ladies talk (and I have my tea, next to me, haha), but it is interesting to see how nVidia is more and more being left behind. I do believe they'll have to go all out with Maxwell's successor. I believe that thing will be able to be something like an APU, but using ARM ISAs, right?

--
@de5_Roy: well, as subtle of a jab as it was, I do believe we'll get even better products, right? haha

Cheers!

 

[jimmysmitty]

Some rumors say Microsoft has already ported Windows server to ARM and engineers are playing with it behind the doors.

Have some links to share about that? I could ask the tech guys in our server division, but if they're under NDA, I really doubt they'll tell me anything, lol.

Mainly because NVidia was not able to get a x86 license. Do you remember the debacle when Intel went from FSB to IMC? NVidia didn't have a license that covered anything beyond Intels FSB design for chipsets. So NVidia lost their ability to make chipsets. Thank God as their chipsets were just bad.

After that I am sure NVidia wanted to branch out and would have killed for a x86 license. I am sure they still would.

I agree with you there, but nVidia has been very stupid about some key things when dealing with Intel. I remember reading that Mr. Huang, when there was talk about a possible Intel buy-out/merger with nVidia, they were very arrogant. I'm still wondering if all that was true or not, but I've always been under the impression that nVidia is sitting on their own little island because they, as an organization, are very stubborn and proud, but in a bad way.

Oh well, this doesn't get more than just old ladies talk (and I have my tea, next to me, haha), but it is interesting to see how nVidia is more and more being left behind. I do believe they'll have to go all out with Maxwell's successor. I believe that thing will be able to be something like an APU, but using ARM ISAs, right?

--
@de5_Roy: well, as subtle of a jab as it was, I do believe we'll get even better products, right? haha

Cheers!

NVidia always has been like that. Always think they are the best but ATI always had a way to shut them down. AMD, not as much but has been decent at it.

 

[anxiousinfusion]

Mainly because NVidia was not able to get a x86 license. Do you remember the debacle when Intel went from FSB to IMC? NVidia didn't have a license that covered anything beyond Intels FSB design for chipsets. So NVidia lost their ability to make chipsets. Thank God as their chipsets were just bad.

After that I am sure NVidia wanted to branch out and would have killed for a x86 license. I am sure they still would.

Is there a reason they haven't approached VIA?

 

[bmacsys]

Yes and it was slower than anything now image that on a server controlling the login process of 500 computers.

Well, if you or your boss are thinking about using a Tablet ARM CPU/SoC to run a server farm, well... You guys need to revisit your ideas.

Don't mix the topics, since they're very different.

Windows RT and software "portability" is one topic on its own, and how the ARMv7 and ARMv8 with their consequent CPU uAarchs are able to handle themselves against Intel and AMD in the vertical markets is another topic.

The PS4 architect tells the anecdote that during the hardware development phase one prominent boss at Sony shared his disagreement by the choice of x86 architecture, because this boss claimed that x86 was an inadequate architecture for gaming. I suppose that one can find ignorant bosses everywhere.

I mentioned before that Microsoft has participated in the development of the new SBSA (Server Base System Architecture) used in AMD Seattle SoC. It seems this important point was ignored.

Those the words of Microsoft representative about SBSA:

http://www.computerworld.com/s/article/9245854/Microsoft_joins_ARM_server_effort

We can confirm Microsoft participates in an industry forum driven by ARM around the Server Base System Architecture, as a part of our continuous work with industry partners to deliver solutions our customers want

Some rumors say Microsoft has already ported Windows server to ARM and engineers are playing with it behind the doors.

Of course they have a port of Windows Server running on ARM. Just like Apple had OSX running on x86 since day one in the skunk works. These are multi-billion dollar corporations not somebody in his garage. They have to be ready for any eventuality.

 

[de5_Roy]

the collective obsession over juanrga is off-topic.

i wanted to discuss about arm and x86 cores being swappable on the same platform. from the s/a articles, it looks like the biggest obstacle is memory access i.e. how an x86 core and arm v8 core would access memory. do they mean cache access or system memory? i assumed both arm and x86 cores in project skybridge will come with their own L2 caches and they will need some kind of hardware virtualization to make the socket compatibility real and software virtualization to make o.s. compatibility real. anyone know how that is possible? is it possible without virtualization? i mean, an always-on, real time hardware and software virtualization seem expensive in terms of power use.
another issue is with the igpu. not the performance, the configuration. current beema soc has 128 radeon cores and 4 a.c.e. imo, amd will need much more if they want to have some relevance in the entrly level market in the future. they'll need at least triple the shader count and 8 a.c.e. (stripped down spectre igpu, to be close) and improved internal cache system for full h.s.a. support.
i don't expect (but will be happy to see) price-range-defying performance from project skybridge since the target usages don't seem performance-hungry. at all.

 

[juanrga]

The only comparison to Haswell I see in that article is to the 192 micro-ops that both Haswell and Apples Cyclone have and the same branch mispredict. Other than that it is more of a comparison to Swift.

There are more comparisons to Intel big core architecture in the article, like this:

With six decoders and nine ports to execution units, Cyclone is big. As I mentioned before, it's bigger than anything else that goes in a phone. Apple didn't build a Krait/Silvermont competitor, it built something much closer to Intel's big cores. At the launch of the iPhone 5s, Apple referred to the A7 as being "desktop class" - it turns out that wasn't an exaggeration.

It is worth mentioning that Piledriver is only a 4-wide architecture.

And no one has said anything about comparing Haswell to ARM. I have already said Haswell is a much more complex chip with superior IPC and will eat it alive.

Read above Anand quote again. Cyclone is already close to Haswell in architectural complexity and IPC. It is "desktop class".

Once ARM starts to throw in that same level of complexity as x86 to achieve the same performance on all levels, that's when the ARM TDP advantage will go away.

ARM efficiency advantage remains at higher TDPs by reasons explained before. Note: Nvidia blog linked before mentions some reasons why ARM is more efficient than x86 and Keller mentioned the efficiency advantage of ARM over x86 during last conference.

I like the part talking about the more secure free alternatives. I hope they realize that the more popular those free alternatives get the more unsecure they become. Look at Apple. Mac OSX has had a big jump in viruses. Android, a Linux based OS, has multiple viruses now and multiple anti-virus applications. It is the nature of the beast. If Microsoft becomes smaller it will just become more secure as the hackers and virus makers will move on to the bigger target.

Let me ask you. Is t more profitable for the Cryptolocker virus to attack Mac OSX, Linux or Windows?

I also like that part, because he is simply repeating that everyone knows: "linux is much more secure than windows"

From Linux The Clear Choice For Security:

Recently, the United Kingdom's Communications-Electronics Security Group (CESG) ran a series of tests to find out which operating system would be the most secure platform for the UK government.

[...]

The goal was to see which platform would pass most of the 12 tests. The winner, [linux], was far ahead of both Windows 8 and Mac OS X.

Your question is easy to answer. Linux has much more marketshare than windows (check tomshardware article), a priori attacking the more popular system is better. However, linux is very secure, it is much more easy for hackers to attack something as insecure and vulnerable as Windows. Windows is a much more lucrative target because it is easier to break and get benefits.

Mainly because NVidia was not able to get a x86 license.

Licensing is another aspect where ARM is superior to x86. However, the Nvidia blog focus was on the technical merits of ARM.

 

[juanrga]

Have some links to share about that? I could ask the tech guys in our server division, but if they're under NDA, I really doubt they'll tell me anything, lol.

No. Sorry.

 

[juanrga]

After Anand analysis showed that Cyclone architecture is "desktop class" and can compete with big Intel cores, an Intel fan did run a comparison between cyclone core and sandy bridge core and got that both are comparable on IPC terms.

http://forums.anandtech.com/showpost.php?p=35511711&postcount=1

The results are unsurprising because Cyclone is already ahead of Piledriver:

Cyclone IPC > A57 IPC
A57 IPC > Jaguar IPC
Jaguar IPC > Piledriver IPC

 

[Lessthannil]

Looks like i'll bite on the AMD A10-7850K since I can get it at Microcenter for $129.

 

[8350rocks]

AMD held APU14 today in China. According to this talk, AMD plans to not move away from desktop. Rather, talks even about growth. Who here was saying that AMD plans to abandon desktop?

I have it on VERY good authority that a new HEDT platform is in the works...my latest information says it will be a little way out, but will offer dCPU and APU options...

Unfortunately, I cannot share more than that...be PATIENT.

(Just FYI for those who would want to know: ARM is not expected to take a significant portion of desktop marketshare...)

 

[blackkstar]

ARM will win in HPC and Servers. It won't unseat x86 desktop and laptops. ARM will win only in places where it's transparent to the user and they have no idea what chip they are running. Meaning situations like Android and iOS where there is an ecosystem of applications to satisfy end users to the point where they can not worry about not having their Windows or x86 applications. Or in situations like a web server where no one cares what CPU is serving them a web page or their files.

You will notice that some people have fallen for the logical fallacy of seeing AMD talk about ARM and x86, AMD saying ARM will win, and then assuming that AMD is only talking about the fight between x86 and ARM and that ARM will unseat x86.

As you can see from the AMD slide referencing projected market shares, ARM is planning on taking a massive chunk away from the "other" category and not the x86 one.

Consumer ARM device sales also depend heavily on a much faster upgrade cycle than a typical x86 product. Most people keep their laptops and desktops until they are too slow. Phones get replaced much more often because new phones have traditionally offer a better user experience that came at a premium price.

http://www.wired.com/2014/05/cheap-smartphones

That is about to change. You are about to see situations where people can buy really cheap phones that work just as well as higher end ones while being able to use those phones for much longer.

I question you all to think back to the Android 2.0 days and such and how big the next hardware release was for your phone. It was huge. It changed the user experience and made things a lot smoother.

Now, we are reaching the point with *CONSUMER* ARM where we are having to resort to benchmarks to prove a point that something is better because it's not so blatantly better than people will upgrade without thinking about.

The rapidly growing consumer ARM market is about to come crashing down to x86 levels and the boom is about to end. People thinking that ARM will continue this growth in the *CONSUMER* space at the cost of x86 market share are very confused as to what is actually happening.

But back to AMD's projected x86 market share numbers. You'll see that after ARM gets done ravaging the other section, it levels off. The graph is awfully deceitful because it starts at 2005. But if you look at what AMD sees happening to x86 marketshare between 2013 and 2018, it's barely going anywhere. In fact, ARM isn't going anywhere either. Also, the ARM and x86 graphs are both taking up roughly the same amount of space. Meaning that between ARm and x86 growth levelling off and AMD projecting ARM and x86 to level off to nearly perfectly split market share, tells me that AMD is not going to abandon x86 like a lot of you are saying.

 

[colinp]

AMD held APU14 today in China. According to this talk, AMD plans to not move away from desktop. Rather, talks even about growth. Who here was saying that AMD plans to abandon desktop?

I have it on VERY good authority that a new HEDT platform is in the works...my latest information says it will be a little way out, but will offer dCPU and APU options...

Unfortunately, I cannot share more than that...be PATIENT.

(Just FYI for those who would want to know: ARM is not expected to take a significant portion of desktop marketshare...)

Quick question, open to all. Isn't a normal feature of an NDA that you can't even reveal that you are under an NDA, never mind give some not-so-subtle hints on what the NDA is about?

 

[8350rocks]

AMD held APU14 today in China. According to this talk, AMD plans to not move away from desktop. Rather, talks even about growth. Who here was saying that AMD plans to abandon desktop?

I have it on VERY good authority that a new HEDT platform is in the works...my latest information says it will be a little way out, but will offer dCPU and APU options...

Unfortunately, I cannot share more than that...be PATIENT.

(Just FYI for those who would want to know: ARM is not expected to take a significant portion of desktop marketshare...)

Quick question, open to all. Isn't a normal feature of an NDA that you can't even reveal that you are under an NDA, never mind give some not-so-subtle hints on what the NDA is about?

That depends on the NDA...for example, I can talk about things that are not proprietary/trade secret/unrevealed items without descriptions or specifics, only to say that something along these lines is coming...

Meanwhile, if you asked me about the project I am working on currently for the studio I work for, I can tell you that I am on a team, on a project. That's about it...because of the sensitivity of it.

Everyone knows AMD makes CPUs and GPUs and that is what they do. To say that something along those lines is coming is very vague at best, however, I can say that I know it is coming...instead of others speculating that it may or may not.

Now, let juanrga talk about Wintelarm servers some more, and about how a coprocessor is going to run an entire server system without a central logic unit to power it...

 

[juanrga]

ARM will win in HPC and Servers. It won't unseat x86 desktop and laptops.

More popular laptop at Amazon is powered by ARM.

You will notice that some people have fallen for the logical fallacy of seeing AMD talk about ARM and x86, AMD saying ARM will win, and then assuming that AMD is only talking about the fight between x86 and ARM and that ARM will unseat x86.

As you can see from the AMD slide referencing projected market shares, ARM is planning on taking a massive chunk away from the "other" category and not the x86 one.

x86 --> [ x86 + ARM ] --> ARM

AMD is now in ^^^ second stage of the transition.

 

[juanrga]

I have it on VERY good authority that a new HEDT platform is in the works...my latest information says it will be a little way out, but will offer dCPU and APU options...

Everyone knows that. Since the 2014 desktop roadmap was released, we know that AMD is extending the Piledriver FX CPU series up to 2015, before has ready the replacement for the series.

What about mentioning details of the replacement? What about the new FX series including 8/12/16 cores? Or about the new FX branding including APUs as well? Or about offering HBM as option? Or about the new FX series abandoning the CMT architecture? Or about the new cores being tiny and efficient? Or about being made on 14nm FinFET...

This is the kind of information expected in an AMD CPU speculation thread.

 

[gamerk316]

Mainly because NVidia was not able to get a x86 license. Do you remember the debacle when Intel went from FSB to IMC? NVidia didn't have a license that covered anything beyond Intels FSB design for chipsets. So NVidia lost their ability to make chipsets. Thank God as their chipsets were just bad.

After that I am sure NVidia wanted to branch out and would have killed for a x86 license. I am sure they still would.

Is there a reason they haven't approached VIA?

Via's license is non-transferable, so NIVIDIA can't get in that way.