[Austin Statesman] AMD sees a way forward (with new Zen design)

[ShintaiDK]

40% may come from something entirely different. It may not even be true at all. Who knows.

Phenom was 40-50% faster than Core 2.
Bulldozer would increase IPC and perform with the best.

The last time AMD didn't flat out lie about performance of a completely new uarch was K8. And that's 12 years ago.

 

[Dresdenboy]

Okay. So AMD claims that Zen will have a 40% increase in IPC over Excavator. It is generally assumed by pessimists that the 40% number refers primarily to FPU-heavy workloads (Cinebench R10 being an example) where Construction cores are considered to be weak. The general idea is that Zen will have more execution resources per core versus BD/PD/SR/XV's modules.

If somebody could show you Cinebench R10 numbers for Carrizo (no throttling, no turbo, just steady clocks), and then we multiplied those numbers by 1.4, would you accept that as a probable performance reference point for Zen?

Carrizo got some nice FP improvements vs. SR (according to Aida64 instruction measurements) possibly at the cost of increased power consumption, so getting throttled back. E.g. sqrt and div latencies are at Intel levels now. There are more execution efficiency improvements. So I think they might even use just a slightly improved FPU for Zen, maybe with more flexible ports.

FPU and integer performance doesn't only depend on raw instruction throughput, but also on data availability (caches, prefetchers), OoO LS efficiency, front end.

 

[Dresdenboy]

40% may come from something entirely different. It may not even be true at all. Who knows.

Indeed the reasons can be many. The memory subsystem is a candidate as they already mentioned caches. And besides instruction throughput Intels bigger cores are known for high cache t'put and low latencies incl. low inter core latencies.

Phenom was 40-50% faster than Core 2.
Bulldozer would increase IPC and perform with the best.

The last time AMD didn't flat out lie about performance of a completely new uarch was K8. And that's 12 years ago.

Are there any proofs for your Phenom performance accusations? Randy Allen talked about Barcelona vs. Clovertown.

 

[CHADBOGA]

Randy Allen talked about Barcelona vs. Clovertown.

So many people forget that before John Freuhe, there was Randy Allen.

 

[VR Enthusiast]

That was the 2000 Intel, but look at the 2013 Intel, that priced Bay Trail low enough to wipe out AMD cat line from the market.

While simultaneously wiping out their CapEX due to massive, recurring mobile losses. Few commentators will say Intel's mobile moves make any kind of sense.

Niche? You haven't been reading market reports in a while. Atom is bigger than AMD was three years ago.

When it comes to losing money I doubt anything is bigger than Atom, not even AMD

 

[mrmt]

While simultaneously wiping out their CapEX due to massive, recurring mobile losses. Few commentators will say Intel's mobile moves make any kind of sense.

Not sure what you are trying to say with this post here.

What I was trying to illustrate is that Intel is pursuing a different strategy than in the 2000's with P4/Centrino and K8 and the market was booming. Intel didn't put Bay Trail in the same margin levels of Core, but instead decided to place it low enough to wipe AMD out of the PC bottom market but still making a healthy amount of money out of it.

That it failed to crack into the mobile market has nothing to do with this strategy being executed, it could even be deepened if Intel wasn't having to sustain the contra-revenue.

So if Fjodor was expecting Intel to keep prices and margins where they are and let AMD choose how to better crack at Core market, better for him to not hold his breath, as the PC market and Intel are very different today. In a depressing marketing like this I don't see Intel conceding one iota to AMD without putting a fight, and given AMD financial situation, that's not a fight they can sustain for too long.

 

[AtenRa]

Not sure what you are trying to say with this post here.

What I was trying to illustrate is that Intel is pursuing a different strategy than in the 2000's with P4/Centrino and K8 and the market was booming. Intel didn't put Bay Trail in the same margin levels of Core, but instead decided to place it low enough to wipe AMD out of the PC bottom market but still making a healthy amount of money out of it.

That it failed to crack into the mobile market has nothing to do with this strategy being executed, it could even be deepened if Intel wasn't having to sustain the contra-revenue.

Who do you trying to fool with that ??? ATOM is not making any money for Intel, on the contrary is 4B losses per year.

Intel didnt decided to push AMD out of the PC market with the Contra-Revenue, they did it in order to gain Tablet market share out of the ARM players. AMD was the unlucky one to be caught in the fight between Intel vs ARM in the tablets at the worst possible time.

ps. The moment Intel stops the Contra-Revenue you will see the ATOM volume sales to plummet like no tomorrow.

 

[VR Enthusiast]

Not sure what you are trying to say with this post here.

What I was trying to illustrate is that Intel is pursuing a different strategy than in the 2000's with P4/Centrino and K8 and the market was booming. Intel didn't put Bay Trail in the same margin levels of Core, but instead decided to place it low enough to wipe AMD out of the PC bottom market but still making a healthy amount of money out of it.

Intel made a healthy amount of money out of it? Does this include or not include contra revenue?

That it failed to crack into the mobile market has nothing to do with this strategy being executed, it could even be deepened if Intel wasn't having to sustain the contra-revenue.

So if Fjodor was expecting Intel to keep prices and margins where they are and let AMD choose how to better crack at Core market, better for him to not hold his breath, as the PC market and Intel are very different today. In a depressing marketing like this I don't see Intel conceding one iota to AMD without putting a fight, and given AMD financial situation, that's not a fight they can sustain for too long.

I think AMD already conceded the ultra-low margin market years ago. It makes sense to do so when faced with much larger companies who are willing to lose billions of dollars fighting over the ultra-low end market.

I think that Intel wiping AMD out of that market is the best thing they could have done for AMD. Now AMD is focused on the high margin chips for 2016 and beyond, while still making guaranteed money on the low margin consoles.

If that was Intel's intention then it has surely backfired as even TSMC now spends much more on CapEX and even GlobalFoundries is catching up. Driving low-margin money out of AMD and in to both of those foundries while at the same time costing billions in contra revenue seems like the kind of plan that has led Intel to their current delays.

 

[mrmt]

Intel made a healthy amount of money out of it? Does this include or not include contra revenue?

Intel already said that on the PC market they are making money with Atom. Atom only loses money when we take into account the tablet market.

I think AMD already conceded the ultra-low margin market years ago. It makes sense to do so when faced with much larger companies who are willing to lose billions of dollars fighting over the ultra-low end market.

They hadn't, really. Did you notice a serious deterioration on AMD CPU business since Q313? That's because Bulldozer wasn't really making money since launch, whatever positive cash flows came from Brazos and then Jaguar. And AMD expected to be able to continue rolling out chips for this segment because they never froze releases for it, even if it was small tweaks like Mullins. That changed when bay trail came and wiped out the cat family, that, coupled with Kaveri having far worse market performance than expected and Maxwell eating GCN for breakfast in terms of market share forced another two rounds of cuts and the fall of the CEO.

I think that Intel wiping AMD out of that market is the best thing they could have done for AMD. Now AMD is focused on the high margin chips for 2016 and beyond, while still making guaranteed money on the low margin consoles.

This is a curious statement. AMD has been trying for... 10 years to develop a high-end chip, and Bulldozer was not as resource constrained as Zen is being. What makes you think that by choking AMD cash flows to unsustainable levels was a good thing for AMD or Zen? And what guarantee does anyone have that Zen will somehow succeed where other much more resourceful, successful and smarter companies like IBM and SUN have failed?

If that was Intel's intention then it has surely backfired as even TSMC now spends much more on CapEX...

Why shouldn't TSMC spend? TSMC CAPEX was a reaction for their market growing and fab costs spiraling, not to anything Intel.

and even GlobalFoundries is catching up.

You have a very distorted vision of the market. Globalfoundries >>failed<< in develop both their 20nm node and their finfet node, they had to license technology from Samsung and doesn't seem to have the money to even buy the same equipment, they are adapting the node, not copying as they should. If anything I'm smelling yet another snafu for 2016, with Globalfoundries unable to deliver and customers not touching them without a 20 feet pole, except for AMD.

 

[frozentundra123456]

Intel made a healthy amount of money out of it? Does this include or not include contra revenue?



I think AMD already conceded the ultra-low margin market years ago. It makes sense to do so when faced with much larger companies who are willing to lose billions of dollars fighting over the ultra-low end market.

I think that Intel wiping AMD out of that market is the best thing they could have done for AMD. Now AMD is focused on the high margin chips for 2016 and beyond, while still making guaranteed money on the low margin consoles.

If that was Intel's intention then it has surely backfired as even TSMC now spends much more on CapEX and even GlobalFoundries is catching up. Driving low-margin money out of AMD and in to both of those foundries while at the same time costing billions in contra revenue seems like the kind of plan that has led Intel to their current delays.

So you cant compete in the high end performance wise and "concede" the low end? Great strategy. Your strategy sounds like a team in the NFL losing as many regular season games as possibe, so as to "focus" on the super bowl. Except your team doesnt have the players to win that either(i.e. the high end).

Edit: sarcasm aside, I dont think AMD conceded anything. They lost the low end for several reasons: decent performance, but still use a lot of power (relatively), poor developer relations, and yes, everybody's favorite excuse for AMD--contra revenue. I mean, look at Carizzo. You cant blame contra revenue for its lack of OEM products. AMD simply either is very poor at it or simply doesnt have the resources to drive OEMs to bring attractive products to the market.

 

[Boze]

Okay. So AMD claims that Zen will have a 40% increase in IPC over Excavator. It is generally assumed by pessimists that the 40% number refers primarily to FPU-heavy workloads (Cinebench R10 being an example) where Construction cores are considered to be weak. The general idea is that Zen will have more execution resources per core versus BD/PD/SR/XV's modules.

If somebody could show you Cinebench R10 numbers for Carrizo (no throttling, no turbo, just steady clocks), and then we multiplied those numbers by 1.4, would you accept that as a probable performance reference point for Zen?

You're really latching onto Cinebench I see, even after I said I threw it out there for the hell of it.

Let me just be as clear as I can be so all the AMD supporters and fanboys get it, for my own personal consumption, then I'll tell you how the mind of every gamer I know or grew up with works.

For me personally:

Whatever AMD releases has to be at least equal to whatever Intel can offer me at the time in terms of performance and, within reason, power consumption. For instance, I recently went to AnandTech's Bench and pulled up the top AMD chip, FX-9590 or whatever it is, and matched it up against my processor, a Core i7 5960X.

It literally dominated every single benchmark. There wasn't a single benchmark that AMD was able to win. Not one. And before everyone goes off the rails crying, you can pull up a Core i7 3960X and see the same thing. Two generations back, Intel's top processor was destroying AMD's best effort. On top of this the FX-9590 was around $800 at release, putting it fairly close to competing against a XX60X processor. But I'll be generous and let you dial it down to a xx30K. And you'll still lose.

For every gamer I know:

Most of the people I know who game are either people I grew up gaming with, playing FPSes or World of Warcraft or EverQuest, and they all have careers and families now. They, like me, care about the best performance they can get. Great example, husband and wife couple I grew up playing EQ with built themselves two new machines for playing Final Fantasy XIV ARR. They used i7-4790K processors and GTX 980s. No AMD in sight.

Some of the younger gamers I know that play DOTA 2 / League of Legends live with Mom & Dad, or are in college with expenses paid for by Mom & Dad, and guess what... Mom & Dad don't buy AMD processors. They buy Intel. They buy Intel because Johnny Knownothing at Best Buy tells them to get this ASUS ROG desktop for Little Timmy, because it'll be able to handle all his school work and he can play a few games because its got a GTX 970 in there.

AMD has to overcome so many obstacles to turn their company around, you'd think they're the reincarnation of Rosa Parks and Martin Luther King, Jr.

They need to win on performance.
They need to win on power consumption.
They need to win on marketing.

It would be great if AMD could manage even one of those... release a 20% slower than Intel's top-end parts, but consume 30% less power. That'd be a home run. Release a part that performs within 5-8% of Intel's top-end parts at the same power consumption. That'd be great. Marketing isn't an option because they don't have the cash to compete against Intel.

Finally, everyone trying to hem and haw and massage whatever numbers they can to make Zen look attractive, you're wasting your time, because all of you consistently try to compare Zen against what Intel offers right now.

Zen is a year away. If you think Intel will sit on the sidelines and let AMD catch up, you are sorely mistaken.

 

[VR Enthusiast]

Intel already said that on the PC market they are making money with Atom. Atom only loses money when we take into account the tablet market.

Right. And does this money they make in the PC market amount to anything useful or is it a drop in the massive 4 billion ocean of losses their mobile dept makes for the company every year?

Any money Intel makes from Atom on the desktop is likely due to economies of scale - economies of scale that require them to lose billions of dollars in mobile in order to "make money" in PC.

They hadn't, really. Did you notice a serious deterioration on AMD CPU business since Q313? That's because Bulldozer wasn't really making money since launch, whatever positive cash flows came from Brazos and then Jaguar. And AMD expected to be able to continue rolling out chips for this segment because they never froze releases for it, even if it was small tweaks like Mullins. That changed when bay trail came and wiped out the cat family, that, coupled with Kaveri having far worse market performance than expected and Maxwell eating GCN for breakfast in terms of market share forced another two rounds of cuts and the fall of the CEO.

Or, there is no point in trying to compete with much bigger companies who are forced to eat massive losses in order to "make money" in razor thin margin PC territory.

This is a curious statement. AMD has been trying for... 10 years to develop a high-end chip, and Bulldozer was not as resource constrained as Zen is being. What makes you think that by choking AMD cash flows to unsustainable levels was a good thing for AMD or Zen? And what guarantee does anyone have that Zen will somehow succeed where other much more resourceful, successful and smarter companies like IBM and SUN have failed?

You seem to have a lot of theories about Zen's R&D but I noticed you rarely back these theories with evidence. So I'll ask again.

Do you have evidence that Zen is "resource constrained" compared to Bulldozer? Any evidence that AMD cut back on Zen R&D?

You have a very distorted vision of the market. Globalfoundries >>failed<< in develop both their 20nm node and their finfet node, they had to license technology from Samsung and doesn't seem to have the money to even buy the same equipment, they are adapting the node, not copying as they should. If anything I'm smelling yet another snafu for 2016, with Globalfoundries unable to deliver and customers not touching them without a 20 feet pole, except for AMD.

The point is that Globalfoundries seems able to "fail" as you put it yet still continue to spend more while Intel spends less. This is true even though oil prices are collapsing and Global is supposed to be dependent on that. You have to wonder what will happen when GlobalFoundries stops failing.

Read this if you haven't already - http://www.kitguru.net/components/a...cts-using-second-gen-14nm-process-technology/

So you cant compete in the high end performance wise and "concede" the low end? Great strategy. Your strategy sounds like a team in the NFL losing as many regular season games as possibe, so as to "fiocus" on the super bowl. Except your team doesnt have the players to win that either(i.e. the high end).

Yes it's called focusing and it's what Apple did in the late 90's. A low-end CPU design on 14nm can cost $200 million or more. That market is long since heading towards the ARM crowd and Intel will be forced to drop out of it as they are forced to continue slashing CapEX to "compete". They can only remain there due to economies of scale which a very long time ago ceased being in their favour.

Anyone who thinks AMD is relevant in this stuff is sadly wrong. Where AMD can make a difference is in high end x86 and servers, and GPU, with the right designs. Intel can't "contra revenue" AMD out of there because Intel needs the high-end profit to survive.

 

[mrmt]

Right. And does this money they make in the PC market amount to anything useful or is it a drop in the massive 4 billion ocean of losses their mobile dept makes for the company every year?

And what makes you think those 4 billions are Atom? Have you been reading the financial statements in Greek?

Any money Intel makes from Atom on the desktop is likely due to economies of scale - economies of scale that require them to lose billions of dollars in mobile in order to "make money" in PC.

It seems that you are mixing OPEX on Intel mobile business with the contra-revenue. If you purge R&D of future mobile components (modems and accelerators included) then Atom makes money.

The point is that Globalfoundries seems able to "fail" as you put it yet still continue to spend more while Intel spends less. This is true even though oil prices are collapsing and Global is supposed to be dependent on that. You have to wonder what will happen when GlobalFoundries stops failing.

If they ever stop failing. Globalfoundries might be competition to UMC or SMIC, but TSMC, Intel or Samsung certainly are concerned by them. They can't develop a node of their own and it seems that they can't implement correctly too. So until their fix their game up they will not be a concern for the guys leading the pack.

But you didn't answer my question, where's GLF catching up? Spending more money on that black hole isn't equal to get results, and right now they are not getting results.

 

[Vortex6700]

AMD simply either is very poor at it or simply doesnt have the resources to drive OEMs to bring attractive products to the market.

...or maybe Intel never stopped doing the things that have made it the most money over the years: Shady business tactics.

Seriously, have we all forgotten that Intel threatened and stopped shipments to OEMs that offered anything AMD?

Not to mention PAYING companies outright to not sell AMD.

How about when they sued Cyrix for 7 years just to save face?

Let's face it, Intel killed AMD when they created a trust with the OEMs. If AMD went away, maybe we would finally see intel get what it deserves, the Ma Bell treatment.

 

[mrmt]

...or maybe Intel never stopped doing the things that have made it the most money over the years: Shady business tactics.

And if this is true maybe AMD BoD reached new levels of incompetence and decided that they will not sue Intel this time.

 

[VR Enthusiast]

And what makes you think those 4 billions are Atom? Have you been reading the financial statements in Greek?

As you are the financial wizard here, why don't you tell us exactly how much Intel makes on desktop atom and how much they lose on atom in tablets?

It seems that you are mixing OPEX on Intel mobile business with the contra-revenue. If you purge R&D of future mobile components (modems and accelerators included) then Atom makes money.

If, but, etc. How much money does Intel make on Atom on PC and how much money does Intel lose on Atom for mobile?

If they ever stop failing. Globalfoundries might be competition to UMC or SMIC, but TSMC, Intel or Samsung certainly are concerned by them. They can't develop a node of their own and it seems that they can't implement correctly too. So until their fix their game up they will not be a concern for the guys leading the pack.

If I were TSMC or Samsung I would be much more concerned by a failing company that continues to spend more money compared to a failing company who continue to spend less.

But you didn't answer my question, where's GLF catching up? Spending more money on that black hole isn't equal to get results, and right now they are not getting results.

Are you privy to GlobalFoundries financial situation and design wins as well as Intel's Atom finances and AMD's R&D spending too? If I had that knowledge I'd probably make a lot of money out of it instead of theorizing on forums if I were you.

And you still didn't answer my question from weeks ago - do you have proof of Zen's reduced R&D? You talk about it all the time so you must have proof of it?

 

[Vortex6700]

And if this is true maybe AMD BoD reached new levels of incompetence and decided that they will not sue Intel this time.

Why? Whole lotta difference it made last time (the antitrust suit).

Also, AMD BoD seems to strive to reach new levels of incompetence.

 

[mrmt]

As you are the financial wizard here, why don't you tell us exactly how much Intel makes on desktop atom and how much they lose on atom in tablets?

Go dig through Intel SEC fillings and the Q&As. The info you want you can get from there, I'm not going to spoon feed you.

 

[frozentundra123456]

...or maybe Intel never stopped doing the things that have made it the most money over the years: Shady business tactics.

Seriously, have we all forgotten that Intel threatened and stopped shipments to OEMs that offered anything AMD?

Not to mention PAYING companies outright to not sell AMD.

How about when they sued Cyrix for 7 years just to save face?

Let's face it, Intel killed AMD when they created a trust with the OEMs. If AMD went away, maybe we would finally see intel get what it deserves, the Ma Bell treatment.

Any proof of that? If so, lets see it.

 

[Boze]

What's funny about all of this to me is that Intel can afford to lose $4 billion a year just to prevent AMD from ever having any hope whatsoever of becoming a part of that market.

Look at AMD's financials... https://www.google.com/finance?q=NASDAQ:AMD&fstype=ii&ei=8BkIVrCfGoSmmAH92LqwBg

AMD's revenue was as much money as Intel lost. That, literally, makes me laugh out loud.

Here's what you all forget about Intel vs. AMD right now. Intel was paying OEMs and threatening them because AMD had a competitive (and at one time, superior) product. AMD doesn't have that right now, and hasn't had it for a decade. Intel has no reason to engage in anti-competitive and monopolistic practices because its already the clear winner.

People talking this s**t about Intel losing money, let me just throw this little movie quote at you:

Charles Foster Kane: You're right, I did lose a million dollars last year. I expect to lose a million dollars this year. I expect to lose a million dollars *next* year. You know, Mr. Thatcher, at the rate of a million dollars a year, I'll have to close this place in... 60 years.

Just replace "Charles Foster Kane" with "Intel" and "a million dollars" with "four billion dollars" and "60 years" with "never".

 

[R0H1T]

What's funny about all of this to me is that Intel can afford to lose $4 billion a year just to prevent AMD from ever having any hope whatsoever of becoming a part of that market.

Look at AMD's financials... https://www.google.com/finance?q=NASDAQ:AMD&fstype=ii&ei=8BkIVrCfGoSmmAH92LqwBg

AMD's revenue was as much money as Intel lost. That, literally, makes me laugh out loud.

Here's what you all forget about Intel vs. AMD right now. Intel was paying OEMs and threatening them because AMD had a competitive (and at one time, superior) product. AMD doesn't have that right now, and hasn't had it for a decade. Intel has no reason to engage in anti-competitive and monopolistic practices because its already the clear winner.

People talking this s**t about Intel losing money, let me just throw this little movie quote at you:



Just replace "Charles Foster Kane" with "Intel" and "a million dollars" with "four billion dollars" and "60 years" with "never".

Yeah keep that up & Intel will be bankrupt inside 10yrs or so, also this ~

Tablet Vendors Rush to Catch Intel&#8217;s Bay Trail CPU Discount Before It&#8217;s Over

http://news.softpedia.com/news/tabl...il-cpu-discount-before-it-s-over-492860.shtml

You saw the performance of A9 right, how long before custom cores from Apple permeate their entire desktop/notebook lineup OR worse custom ARM cores (like Mongoose or Denver) invade servers? If past is any indication then no megacorp is safe forever & anyone thinking Intel can keep on throwing billions for fun needs to be laughed at :hmm:

 

[Vortex6700]

Any proof of that? If so, lets see it.

No problem, as these are well known events.

Do you want proof of the well documented lawsuits Intel lost or settled, or would you like me to prove that Intel killed AMD when they did the things they did?

 

[Boze]

Yeah keep that up & Intel will be bankrupt inside 10yrs or so, also this ~
http://news.softpedia.com/news/tabl...il-cpu-discount-before-it-s-over-492860.shtml

No, they won't be, which is why I said "never". If your profit is $13 billion a year, losing $4 billion still makes you profitable. Practically they cannot lose $4 billion a year because the CEO would be ousted eventually.

You saw the performance of A9 right, how long before custom cores from Apple permeate their entire desktop/notebook lineup OR worse custom ARM cores (like Mongoose or Denver) invade servers? If past is any indication then no megacorp is safe forever & anyone thinking Intel can keep on throwing billions for fun needs to be laughed at :hmm:

You're right about the A9X. Apple is now Intel's competitor, and the company they should be worried about. Another 3-5 years of development and an A11X processor might be able to compete with a desktop / high-end laptop part and replace Intel chips in iMacs.

A very real worry for Intel, or at least it should be.

As nice as the A9X is, I suspect when we see Surface Pro 4 14" come out with a Skylake mobile processor, it will destroy the iPad Pro in benchmarks. Of course, raw performance is not where the iPad Pro fails, its in the iOS system itself, since there's no Adobe Photoshop / Illustrator / Premiere / inDesign for iOS, nor is there any other high-end production software.

That might change, of course, but I want to get back on the subject of AMD because I truly believe their very real and imminent demise will be within the next two years.

Intel is no longer Rome at their gates. Now they've got the Romans and the barbarians beating on the gates and threatening to flood in and destroy them.

I would love to see LG or Samsung or even Microsoft buy AMD, go into dormant mode for a few years, and then come out with a processor that just slaughters everyone, but that's probably too much to hope for.

 

[AtenRa]

It would be great if AMD could manage even one of those... release a 20% slower than Intel's top-end parts, but consume 30% less power. That'd be a home run. Release a part that performs within 5-8% of Intel's top-end parts at the same power consumption. That'd be great. Marketing isn't an option because they don't have the cash to compete against Intel.

I will agree to this, AMD doesnt need to be faster to compete with Intel. But they have to make their products available if they want to get sales up.

 

[R0H1T]

No, they won't be, which is why I said "never". If your profit is $13 billion a year, losing $4 billion still makes you profitable. Practically they cannot lose $4 billion a year because the CEO would be ousted eventually.

You're assuming Intel will keep on making the same $13 billion for the next 10yrs, suffice to say that's a prediction I won't back. The only thing for certain is that the traditional computing realm is shrinking, like I said a couple of years back, & that's a secular irreversible trend.

http://www.gartner.com/newsroom/id/3135618

You're right about the A9X. Apple is now Intel's competitor, and the company they should be worried about. Another 3-5 years of development and an A11X processor might be able to compete with a desktop / high-end laptop part and replace Intel chips in iMacs.

A very real worry for Intel, or at least it should be.

As nice as the A9X is, I suspect when we see Surface Pro 4 14" come out with a Skylake mobile processor, it will destroy the iPad Pro in benchmarks. Of course, raw performance is not where the iPad Pro fails, its in the iOS system itself, since there's no Adobe Photoshop / Illustrator / Premiere / inDesign for iOS, nor is there any other high-end production software.

That might change, of course, but I want to get back on the subject of AMD because I truly believe their very real and imminent demise will be within the next two years.

Intel is no longer Rome at their gates. Now they've got the Romans and the barbarians beating on the gates and threatening to flood in and destroy them.

I would love to see LG or Samsung or even Microsoft buy AMD, go into dormant mode for a few years, and then come out with a processor that just slaughters everyone, but that's probably too much to hope for.

Totally agree with the bolded part but I'm keeping my fingers crossed, just in case :biggrin:

The other thing to note is that Intel's still a pawn in the grand scheme of things, IoT is the next wave incoming. ARM is a monster in embedded systems & the long list of cheap ARM chipmakers is virtually endless. Now I won't predict how consumers will react to this wave but one thing is for certain that tech companies are going to throw billions (probably even trillions) to get a foothold in IoT over the coming decades.

Then comes the real downer, custom ARM cores are getting to the point where they can replace Intel Xeons at the low end & we certainly know that virtually every tech behemoth, like Google & FB, are looking to go this route in the future. Depending on how serious they are & how much $$ they're willing to spend, Intel will be in a world of hurt. If anything what Intel's shown us, especially over the last couple of years with their payouts, is that they're desperate to get into new markets because they're ill positioned to take on the next ARM wave & trust me it's definitely coming. Lastly enjoy this piece of rumor :awe:

http://www.gsmarena.com/apples_nextgen_a10_chipset_rumored_to_have_hexacore_cpu-news-14126.php