Discussion: AMD's last hope for survival lies in the Zen CPU architecture

[synphul]

+1 for fallout 3, somewhat of an old adopter here. Just started playing it, rather enjoyable. Definitely not interested in an os 'upgrade' that kills compatibility/functionality.

In terms of ibm, they did sell off their hardware/systems to lenovo however another reason they likely won't fuss with amd, they're already working with samsung on their own 7nm chip. No reason to go with amd, amd is outsourcing to globalfounderies and so is ibm. Ibm may be getting back into the chip game after all, however it's in the working stages right now and moving away from silicon only with this newer chip design.

 

[NightAntilli]

Google doesnt need to invest billions in a company to get 64bit desktop cpus.

But if AMD fails, then some other company should pick them up, right? Like IBM, Amazon, or Microsoft.

IBM will not buy AMD.

In the last 10 years, they shed off their, x86 server business to Lenovo, Printer business to Rioch and Lexmark, and chip manufacturing facilities to GlobalFoundries. IBM has turned into a service oriented business, and is getting away from consumer level hardware. However, they're still keeping their Power systems, Unix, zOS, and Mainframes, for enterprise level buisness.

Sorry to burst that bubble. But I think Microsoft or Apple might pick up AMD instead.

I hope it's anyone but Apple. If Apple takes over AMD we're gonna have two nVidias and two Intels in the market. We need someone pushing open source rather than closed standards, and Apple is not exactly known for its open source. So I'd rather it be Microsoft, even though that's also far from ideal. In fact, Google might be the best one for the industry. Whether they're interested is another story.

 

[Derek Bartron]

Lack of vision (like avoiding the mobile market in a shrinking pc market) and the lack of innovating beyond just CPU speed and Cores is why AMD loses.

For instance, it wanted to combine the CPU and GPU into one processor. Great idea!, especially if it gets to the point you can actually play games on it. However, why not focus on using this CPU in laptops and pads and create a niche market where this synergy is needed?

Instead as people move to pads and phones for computing work, AMD still tries to beat out Intel and Nvidia while it's total customer base shrinks.

Someday we won't have computers on a desk, they will be carried in our pockets, and avoiding this eventuality and not pushing toward this reality is why AMD is losing.

If AMD rebranded itself as the world of innovation, and got out of the speed game and got into the "lets make it mobile game", you would see them spring to life.

Want proof? Look at how well ARM is doing.

 

[Supernova1138]

Lack of vision (like avoiding the mobile market in a shrinking pc market) and the lack of innovating beyond just CPU speed and Cores is why AMD loses.

For instance, it wanted to combine the CPU and GPU into one processor. Great idea!, especially if it gets to the point you can actually play games on it. However, why not focus on using this CPU in laptops and pads and create a niche market where this synergy is needed?

Instead as people move to pads and phones for computing work, AMD still tries to beat out Intel and Nvidia while it's total customer base shrinks.

Someday we won't have computers on a desk, they will be carried in our pockets, and avoiding this eventuality and not pushing toward this reality is why AMD is losing.

If AMD rebranded itself as the world of innovation, and got out of the speed game and got into the "lets make it mobile game", you would see them spring to life.

Want proof? Look at how well ARM is doing.

Maybe if AMD did that about 5 or 6 years ago they might have been successful, now the mobile market is way too crowded, major players have pretty much entrenched themselves, and even Intel with its large stockpiles of cash struggles to muscle their way into it. AMD doesn't have the cash to put into R&D to get their power consumption down to Intel levels to make their x86 chips more attractive for mobile devices, and they certainly don't have the cash to sell their products at a loss just to try to get some market share in the mobile space like what Intel has been doing. AMD is starting work on an ARM processor, probably so they will be more attractive to buyers as the x86 license is non transferable and AMD needs to make something else, especially if they do wind up selling off the GPU division to keep themselves above water for another couple of years.

 

[amdgraphics]

Lack of vision (like avoiding the mobile market in a shrinking pc market) and the lack of innovating beyond just CPU speed and Cores is why AMD loses.

For instance, it wanted to combine the CPU and GPU into one processor. Great idea!, especially if it gets to the point you can actually play games on it. However, why not focus on using this CPU in laptops and pads and create a niche market where this synergy is needed?

Instead as people move to pads and phones for computing work, AMD still tries to beat out Intel and Nvidia while it's total customer base shrinks.

Someday we won't have computers on a desk, they will be carried in our pockets, and avoiding this eventuality and not pushing toward this reality is why AMD is losing.

If AMD rebranded itself as the world of innovation, and got out of the speed game and got into the "lets make it mobile game", you would see them spring to life.

Want proof? Look at how well ARM is doing.

Maybe if AMD did that about 5 or 6 years ago they might have been successful, now the mobile market is way too crowded, major players have pretty much entrenched themselves, and even Intel with its large stockpiles of cash struggles to muscle their way into it. AMD doesn't have the cash to put into R&D to get their power consumption down to Intel levels to make their x86 chips more attractive for mobile devices, and they certainly don't have the cash to sell their products at a loss just to try to get some market share in the mobile space like what Intel has been doing. AMD is starting work on an ARM processor, probably so they will be more attractive to buyers as the x86 license is non transferable and AMD needs to make something else, especially if they do wind up selling off the GPU division to keep themselves above water for another couple of years.

But what could they make? Is Zen the answer? I don't know, but I am hoping Zen will keep them afloat for a while. When is the last time AMD was ahead of Intel? Who would they sell the GPU division to?

 

[Supernova1138]

Lack of vision (like avoiding the mobile market in a shrinking pc market) and the lack of innovating beyond just CPU speed and Cores is why AMD loses.

For instance, it wanted to combine the CPU and GPU into one processor. Great idea!, especially if it gets to the point you can actually play games on it. However, why not focus on using this CPU in laptops and pads and create a niche market where this synergy is needed?

Instead as people move to pads and phones for computing work, AMD still tries to beat out Intel and Nvidia while it's total customer base shrinks.

Someday we won't have computers on a desk, they will be carried in our pockets, and avoiding this eventuality and not pushing toward this reality is why AMD is losing.

If AMD rebranded itself as the world of innovation, and got out of the speed game and got into the "lets make it mobile game", you would see them spring to life.

Want proof? Look at how well ARM is doing.

Maybe if AMD did that about 5 or 6 years ago they might have been successful, now the mobile market is way too crowded, major players have pretty much entrenched themselves, and even Intel with its large stockpiles of cash struggles to muscle their way into it. AMD doesn't have the cash to put into R&D to get their power consumption down to Intel levels to make their x86 chips more attractive for mobile devices, and they certainly don't have the cash to sell their products at a loss just to try to get some market share in the mobile space like what Intel has been doing. AMD is starting work on an ARM processor, probably so they will be more attractive to buyers as the x86 license is non transferable and AMD needs to make something else, especially if they do wind up selling off the GPU division to keep themselves above water for another couple of years.

But what could they make? Is Zen the answer? I don't know, but I am hoping Zen will keep them afloat for a while. When is the last time AMD was ahead of Intel? Who would they sell the GPU division to?

Last time AMD was ahead of Intel was about 10-15 years ago, and that probably had more to do with Intel screwing up with the Pentium 4 and spending about 4 years trying to push higher clockspeeds at all costs, tanking their IPC and upping their power consumption and heat output in the process.

At this point, AMD is probably finished, and the only way they're going to survive is if Intel or Nvidia really screw up and give AMD an opportunity to regain a lot of market share. The GPU division is the only part that will likely live on when someone buys it. As for who will buy it, who knows, probably one of the mobile chip makers who might want better graphics, maybe Apple will buy it so they can go more proprietary with their Macs by making their own CPUs and GPUs and try to shut down the hackintosh scene.

 

[amdgraphics]

Lack of vision (like avoiding the mobile market in a shrinking pc market) and the lack of innovating beyond just CPU speed and Cores is why AMD loses.

For instance, it wanted to combine the CPU and GPU into one processor. Great idea!, especially if it gets to the point you can actually play games on it. However, why not focus on using this CPU in laptops and pads and create a niche market where this synergy is needed?

Instead as people move to pads and phones for computing work, AMD still tries to beat out Intel and Nvidia while it's total customer base shrinks.

Someday we won't have computers on a desk, they will be carried in our pockets, and avoiding this eventuality and not pushing toward this reality is why AMD is losing.

If AMD rebranded itself as the world of innovation, and got out of the speed game and got into the "lets make it mobile game", you would see them spring to life.

Want proof? Look at how well ARM is doing.

Maybe if AMD did that about 5 or 6 years ago they might have been successful, now the mobile market is way too crowded, major players have pretty much entrenched themselves, and even Intel with its large stockpiles of cash struggles to muscle their way into it. AMD doesn't have the cash to put into R&D to get their power consumption down to Intel levels to make their x86 chips more attractive for mobile devices, and they certainly don't have the cash to sell their products at a loss just to try to get some market share in the mobile space like what Intel has been doing. AMD is starting work on an ARM processor, probably so they will be more attractive to buyers as the x86 license is non transferable and AMD needs to make something else, especially if they do wind up selling off the GPU division to keep themselves above water for another couple of years.

But what could they make? Is Zen the answer? I don't know, but I am hoping Zen will keep them afloat for a while. When is the last time AMD was ahead of Intel? Who would they sell the GPU division to?

Last time AMD was ahead of Intel was about 10-15 years ago, and that probably had more to do with Intel screwing up with the Pentium 4 and spending about 4 years trying to push higher clockspeeds at all costs, tanking their IPC and upping their power consumption and heat output in the process.

At this point, AMD is probably finished, and the only way they're going to survive is if Intel or Nvidia really screw up and give AMD an opportunity to regain a lot of market share. The GPU division is the only part that will likely live on when someone buys it. As for who will buy it, who knows, probably one of the mobile chip makers who might want better graphics, maybe Apple will buy it so they can go more proprietary with their Macs by making their own CPUs and GPUs and try to shut down the hackintosh scene.

I don't know about you, but I feel bad for AMD. They are a good company. I want them to be up with Intel. Intel needs the competition.

 

[Supernova1138]

They do, but at this point the only real competition Intel is going to get would be from outside the x86 scene, which would mean losing backwards compatibility with 30 years worth of software if one were to switch to an ARM or PowerPC based system. Intel might not even technically count as a monopoly if AMD went under because there are other CPU architectures that exist, they simply aren't relevant in the desktop computing space. As such, there might not be any sort of anti-trust action against Intel if they wind up as the sole company producing desktop CPUs and we'll just have to deal with it.

 

[amdgraphics]

They do, but at this point the only real competition Intel is going to get would be from outside the x86 scene, which would mean losing backwards compatibility with 30 years worth of software if one were to switch to an ARM or PowerPC based system. Intel might not even technically count as a monopoly if AMD went under because there are other CPU architectures that exist, they simply aren't relevant in the desktop computing space. As such, there might not be any sort of anti-trust action against Intel if they wind up as the sole company producing desktop CPUs and we'll just have to deal with it.

But since Intel would be best at it, they would probably charge more than before. Consider this: UPS and FedEx are rivals, right? They are the biggest shippers for Amazon etc. There is Ontrac, but they a smaller company. What if FedEx went out of business. UPS could charge way more because people would choose UPS over Ontrac. UPS has no real competition in terms of big business like FedEx gave them.

 

[jimmysmitty]

If AMD can deliver on IPC this time around, and manage to sustain their multicore performance in the process, hopefully they can hold out long enough to delve deeper into their RND and make some headway back into the market.

Brand dosent matter in my eyes, its performance and pricing. If one manufacture cant keep up, its bad for pricing.

The problem is that their current multicore performance fails due to the high power draw. That is why they have lost the majority of their server market share which is way more vital then the lower margin consumer market share. The same Core i7 we pay $350 for on desktop tends to fetch a much higher price on the server side.

From all the rumors that have been going around the IPC increase wont even put it on par with Haswell, it will be slightly lower than it, and Zen is set for release in Q4 2016 meaning it will see the Skylake refresh as its competitor then shortly after will have to deal with the die shrink to Skylake negating its catch up to Intel at 14nm.

It is not doom and gloom, but from a business perspective, AMD has been floundering way too long. Too many restructures and not enough output of more competitive products. Their time between product launches is too long and when they do release something that catches up, i.e. Phenom II, it gets jumped by whatever Intel throws out next. BD was worse as when it launched their marketing team threw up slides with cherry picked benchmarks only in areas where it beat older CPUs from Intel.

I don't think AMD is gone, I still like their GPUs although Fury has less than inspired me but I don't feel as if they are going to be around the same way we have known them. I truly think within 5 years unless they pull another K8 out of their arse they will change drastically. Then again K8 was the perfect storm. It was a decent CPU but Intel also slipped drastically with NetBurst which made K8 look even better than it would have if they didn't do NetBurst.

It is truly an interesting case that has to be watched to see the future of it.

They really should have ditched BD after the first iteration and moved back to a core like Zen. If they did that and focused more on that they might have something else right now that would be more competitive than what they have.

Of course I also did state that one downside of AMD dropping their FABs was that they are now at the mercy of the FAB companies. They have to rely on others to get a good process going and even with a good uArch it can be screwed up by a faulty/leaky process node. A good example is with the Hawaii uArch of GPUs. It was originally slated for 20nm but as we saw nothing went to 20nm and stayed on 28nm. If there was a decent 20nm Hawaii might have been better than it was. This means that now AMD needs to build their CPUs around whatever the best process node that is available to them while Intel builds a new uArch on a current, mature process node then shrinks the known good uArch.

 

[turkey3_scratch]

Kind of joining into the conversation late, but anyway I recall the CEO specifically stating Zen will have a 40% IPC increase. What worries me is that Intel is already above 40% of AMD CPU IPC.

Also, quick question. How much money would, say, a single Fury or 980Ti GPU cost to manufacture?

 

[InvalidError]

Kind of joining into the conversation late, but anyway I recall the CEO specifically stating Zen will have a 40% IPC increase. What worries me is that Intel is already above 40% of AMD CPU IPC.

What I would be more worried about are the circumstances under which that 40% IPC increase is achieved: AMD is 30-40% behind on single-threaded IPC but if the bulk of their 40% IPC increase with Zen is from SMT in multi-threaded workloads, this still leaves AMD the better part of 40% behind Intel on single-threaded performance. Since Intel also gains ~30% from HT, AMD's 40% IPC improvement would only narrow the gap by ~10% in multi-threaded workloads.

When you consider SMT and multi-threaded workloads, it is quite easy to imagine how AMD could justify their 40% claim yet fail to gain much ground in moderately threaded code against Intel CPUs for the same total core and thread count.

 

[turkey3_scratch]

Kind of joining into the conversation late, but anyway I recall the CEO specifically stating Zen will have a 40% IPC increase. What worries me is that Intel is already above 40% of AMD CPU IPC.

What I would be more worried about are the circumstances under which that 40% IPC increase is achieved: AMD is 30-40% behind on single-threaded IPC but if the bulk of their 40% IPC increase with Zen is from SMT in multi-threaded workloads, this still leaves AMD the better part of 40% behind Intel on single-threaded performance. Since Intel also gains ~30% from HT, AMD's 40% IPC improvement would only narrow the gap by ~10% in multi-threaded workloads.

When you consider SMT and multi-threaded workloads, it is quite easy to imagine how AMD could justify their 40% claim yet fail to gain much ground in moderately threaded code against Intel CPUs for the same total core and thread count.

I thought IPC was only related to single-core performance, as doesn't each core carry out its own single simple instruction per cycle, multiple of these cycles carrying out more complex instructions?

Does IPC really have a measurement or is it just irrational? Because wouldn't it technically be a decimal since complex instructions would take lots of clocks. It should really be clocks per instruction - it makes more sense that way. Where a lower CPI is better.

And doesn't IPC increase only actually an increase for particular instructions that are made more efficient? Sorry for firing all these Q's.

 

[InvalidError]

I thought IPC was only related to single-core performance, as doesn't each core carry out its own single simple instruction per cycle, multiple of these cycles carrying out more complex instructions?

It has not been that way in over 20 years since the Pentium got introduced unless you count early Atom SoCs. Some architectures like Alpha started doing pipelined superscalar execution even earlier than that. With the Pentium Pro, x86 also acquired out-of-order execution.

Superscalar execution is the ability to start executing multiple instructions per clock, pipelining is the ability to spread computations in multiple sequential stages while still accepting new operands and producing a result on every clock and out-of-order execution is the ability to shuffle instructions in the decode pipeline to maximize utilization of execution units.

Modern CPUs can issue 3-5 instructions per clock to different execution units, each execution unit is pipelined to accept a new instructions on nearly every clock, some execution pipelines need only one cycle to complete an operation while others might take five or more. At any one time, Intel's modern CPUs can have over a dozen instructions in-flight through their execution units on top of the 192 in the scheduler.

Some more complex instructions may get split into multiple microcode instructions during decoding and pass through some of the longer execution pipelines but these will be mixed with whatever else the scheduler is able to cram down the other execution ports along with it.

As for IPC, that is "Instructions Per Clock" and does not distinguish between single-threaded cores and multi-threaded ones unless explicitly stated as such. Technically, it does not distinguish between the number of cores either, so it would also be necessary to explicitly state "Per Core" too to make sure part of the "40% more IPC" is not from whacking an extra core or two on the die.

 

[Joseph DeGarmo]

If the AMD CPU division goes belly up, at least their GPU division will stay afloat. AMD could be a GPU-only manufacturer as ATI was. But if NVIDIA ends up buying out the GPU division, then things will get really interesting.

 

[gamerk316]

The problem, again, remains the debt AMD has to start paying off in three years. Simply put, a company of AMDs size can't sustain it's current debt load, and if parts of the company go under, the remaining parts simply won't bring in the necessary cash needed to stave off bankruptcy. Hence why it's more likely the easier to sell GPU division gets flipped for cash, which is then used to pay off part of their debts.

 

[-Fran-]

The problem, again, remains the debt AMD has to start paying off in three years. Simply put, a company of AMDs size can't sustain it's current debt load, and if parts of the company go under, the remaining parts simply won't bring in the necessary cash needed to stave off bankruptcy. Hence why it's more likely the easier to sell GPU division gets flipped for cash, which is then used to pay off part of their debts.

And that is also a problem.

Selling the GPU division will not bring enough money to fill all the holes the company has. Even more, they would be getting rid of a big portion of income.

The only reason AMD would get rid of the GPU division, is because shareholders want to cash in before AMD really goes bankrupt. I smell a big torment if such a thing would happen.

Cheers!

EDIT: Typos.

 

[gamerk316]

Selling the GPU division will not bring enough money to fill all the holes the company has. Even more, they would be getting rid of a big portion of income.

It's loosing money, and at the end of the day, it's probably the only way AMD can raise the $600 Million in needs to have on hand in 2019.

Course, that implies they sell ATI at 25% of what they brought it at, but AMD probably doesn't have a choice. They're pretty much out of cash, and Zen is their last real chance to turn a sustainable profit. AMD simply can't survive as it's currently constructed.

 

[InvalidError]

Hence why it's more likely the easier to sell GPU division gets flipped for cash, which is then used to pay off part of their debts.

With APUs and SoC chips becoming increasingly important in the future, ditching their graphics IP seems suicidal. It might raise cash to last a few more years but without income from SoCs and APU sales, I doubt what would be left of AMD would have any chance of recovering.

 

[jimmysmitty]

Hence why it's more likely the easier to sell GPU division gets flipped for cash, which is then used to pay off part of their debts.

With APUs and SoC chips becoming increasingly important in the future, ditching their graphics IP seems suicidal. It might raise cash to last a few more years but without income from SoCs and APU sales, I doubt what would be left of AMD would have any chance of recovering.

It is possible to sell it off but keep rights to or work with the company to keep their APU side afloat. Much like Ford who has sold off most of their interest in Mazda yet have kept using a lot of designs they got from them or same with Landrover/Jaguar which still use Bendix parts and such.

Not saying they will but it is a possibility.

Or they could decide to develop their own version based on whatever the current uArch they have for GPUs and go with that while the GPU spin off develops its own. Intel seems to be doing just fine without a dGPU business, in fact their iGPUs keep getting faster and closing in on the mid range dGPU market.

 

[gamerk316]

Hence why it's more likely the easier to sell GPU division gets flipped for cash, which is then used to pay off part of their debts.

With APUs and SoC chips becoming increasingly important in the future, ditching their graphics IP seems suicidal. It might raise cash to last a few more years but without income from SoCs and APU sales, I doubt what would be left of AMD would have any chance of recovering.

I could say the same exact thing about AMD selling their Fabs.

In any case, I'm sure they'd have a robust licensing agreement in place for the foreseeable future, just like they had with GloFo.

 

[InvalidError]

Or they could decide to develop their own version based on whatever the current uArch they have for GPUs and go with that while the GPU spin off develops its own. Intel seems to be doing just fine without a dGPU business, in fact their iGPUs keep getting faster and closing in on the mid range dGPU market.

If I was buying ATI back from AMD, I would not allow AMD to retain any rights to the GPU architectures if they want a remotely decent price: who would buy the remains of a GPU designer if the parent company retains enough rights to arbitrarily spring CPUs with high-end IGPs on you? If they want to continue making CPUs, APUs, SoCs and whatever else that use GCN IP, they will have to license it back from whoever they (hypothetically) sell ATI to.

As for Intel's IGPs doing fine without dGPUs, that is only because the IGP's primary function is to make the die bigger - the regular IGP on Intel's lower-end chips is mostly hopeless. Yes, Intel has IrisPro, but those come with hefty price premiums to deliver barely entry-level gaming graphics, which does not make any sense unless your form factor or power budget cannot accommodate anything else. Also, Intel's current IGP architecture probably would not scale well enough to compete against higher-end discrete GPUs if Intel decided to try offering discrete GPUs again.

 

[Rodrigodrt]

I think its time for qualcomm to delve into x86. AMDone, I wouldnt expect much, but to be honest I wish they would come around. Used to love my Athlon. Used to.

 

[AMDHTPC]

Then if AMD is on its way out, it could eventually become the intellectual property of Google as a result of a buyout. It would make more sense if Google would purchase AMD because it could use the APUs in chromebooks and Android-based smartphones. Even Google could invest in marketing pre-built gaming Chrome desktops and consoles using AMD FX CPUs, RAM, and Radeon GPUs. Google could also release first-party in-house games exclusive to those machines the same way Nintendo does with its consoles.

You know normally I'd dismiss these types of speculation as pure fantastical bullshit, but you know what? I think you might be right on the money with this one. The way Google has been buying up robotics companies I don't think this is as far fetched as one may initially think.

Remember that Google is basically the internet branch of the NSA and an extension of Darpa by owning Boston Dynamics (creators of the famous "BigDog" robot among many others).

Please god don't let Google ruin AMD!

I think AMD will turn this thing around to be at least competitive with Intel rather than lagging behind. It wasn't too long ago the Athlon rocked Intel's world, you know they had to be shocked when the perpetual underdog had beat them. This is why the existence of AMD & Intel is beneficial to all. Competition & innovation are the keys to their relationship and is what gives us, the consumer, lower prices on cutting edge technology. I believe AMD may qualify as a "too big to fail company" that could get bailed out somehow. The fallout from AMD biting the dust would send shock waves throughout the tech industry causing a cascade of tech companies to go bankrupt. The ramifications would be huge. Between consoles, their APU's and the GPU division they should be able to overcome much of the debt, but at the end of the day Zen is going to be a huge factor determining the future of AMD.

 

[jimmysmitty]

Then if AMD is on its way out, it could eventually become the intellectual property of Google as a result of a buyout. It would make more sense if Google would purchase AMD because it could use the APUs in chromebooks and Android-based smartphones. Even Google could invest in marketing pre-built gaming Chrome desktops and consoles using AMD FX CPUs, RAM, and Radeon GPUs. Google could also release first-party in-house games exclusive to those machines the same way Nintendo does with its consoles.

You know normally I'd dismiss these types of speculation as pure fantastical bullshit, but you know what? I think you might be right on the money with this one. The way Google has been buying up robotics companies I don't think this is as far fetched as one may initially think.

Remember that Google is basically the internet branch of the NSA and an extension of Darpa by owning Boston Dynamics (creators of the famous "BigDog" robot among many others).

Please god don't let Google ruin AMD!

I think AMD will turn this thing around to be at least competitive with Intel rather than lagging behind. It wasn't too long ago the Athlon rocked Intel's world, you know they had to be shocked when the perpetual underdog had beat them. This is why the existence of AMD & Intel is beneficial to all. Competition & innovation are the keys to their relationship and is what gives us, the consumer, lower prices on cutting edge technology. I believe AMD may qualify as a "too big to fail company" that could get bailed out somehow. The fallout from AMD biting the dust would send shock waves throughout the tech industry causing a cascade of tech companies to go bankrupt. The ramifications would be huge. Between consoles, their APU's and the GPU division they should be able to overcome much of the debt, but at the end of the day Zen is going to be a huge factor determining the future of AMD.

The issue with any buyout is the transfer of the x86 license. That determines a lot.

And Zen will only be successful if it is a competitive server chip. People keep assuming that the consumer market is where Zen needs to excel but it really does not. AMD used to have 25% of the server market share or more. That started when they first launched the Opteron which dominated performance and power metrics. But since Intel got their head in the game and AMDs server variants have been lack luster they have dropped to less than 5%.

If Zen is competitive in the server market and AMD can pull server design wins they will be able to survive much better than if they secure consumer wins.

What would you rather have. A market where the majority of people buy the cheapest product or a market where the majority buys the best performance/watt product for 3-4x more? Margins are way better in the server and workstation market.