With the current rate of Intel CPU performance increases, could AMD be catching up?

[mrmt]

Of course not. Just like I said, they're in bad shape and in a bad situation. But imho their current market position is better than the one 1-2 years ago. Talking about doom and gloom now strikes me as odd, honestly.

You are assuming that what they have will sell, and that it will sell enough to keep the R&D engine running. The second assumption is far fetched.

 

[ShintaiDK]

That's not a node advantage, that is economy of scale. So what you mention is not any technological advantage of moving to a later node.

As long as there's a company that can produce AMD's wafers they should not be affected by it, since the cost per transistor apparently is about the same on 14 vs 22 nm.

Also note that GloFo doesn't produce wafers for AMD only. They have other customers too like Qualcomm keeping the volumes up.


Your own diagram concluded that the benefit of moving to a newer node was not that great anymore.

Also, AMD is fighting with 28 nm parts (Trinity) vs 22 nm parts (IB) for Intel. If you're comparing Kaveri (28 nm) vs Broadwell (14 nm) I think that is unfair, since Kaveri is expected to be released in late 2013 vs Broadwell in mid-2014. Also, you are comparing at a point where the node difference is at its peak, with Broadwell being a tick meaning that Intel will just have switched over to 14 nm vs Kaveri that likely is at the end of the 28 nm tock cycle for AMD.

Transistor cost is not the same. And Qualcomm is a small company compared to Intel in revenue.

AMD is already losing money, -37% YoY CPU revenue. Their CEO saying its done for with big cores and so on. What more do you need as edvidence?

GloFo is a terrible foundry in tersm of delivery of nodes. They are already way behind and constantly changes their roadmaps to reflect it.

So you say the 6 months that AMD gets on 28nm before Intel releaes 14nm makes a difference? When will AMD hit 20nm? 2015? 2016? 20nm vs 10nm?

And people dont care if poor AMD is behind in processnodes. They will buy a 22 or 14nm product with much better power and performance charateristics. Just as they do today and why AMD is failing so hard in the CPU division.

The only question is, will AMD go belly up or end like VIA.

 

[ShintaiDK]

Isn't it strange how people will on one hand have absolutely no issues accepting the economical realities behind why every single other x86 design house went belly up, withdrew from the market, or shriveled and shrunk to fit within a niche...but it suddenly becomes inconceivable that those same concepts might be applicable to AMD?

This is the part that baffles me. I can understand how non-professional people might not be able to grasp that an Intel operates at economies of scale in R&D that makes its R&D dollars deliver more per dollar than other companies (design house synergy)...after all many of these folks have never seen the inside of a fab, or attended a spec meeting for spice models on a node that won't see a fab for 4 yrs yet.

But what baffles me is that despite all they don't know, they will readily accept the economics of what they think they know as applied to Cyrix, National Semi, Via, Texas Instruments, Transmeta, IDT, etc...but don't you dare try and explain how AMD is not so special as to be excluded from that list, it becomes a sacred cow at that point for some reason. To be defended to the death (or the absurd as I read in the past many posts in this thread).

Via never had it so good, no one ever valiantly said "Via will make it happen, Intel and AMD can't possibly beat them into the ground, look at the cost of wafers that impacts AMD and Intel too.." etc etc.

Nope, for some reason it all made perfectly logical sense that Via (and everyone else in x86) went away as they did...but those same reasons simply can't be applied to AMD. Ever.

Brand loyalty or delusion is the only thing I can think on. The numbers as you say dont lie. Its about equal to expect Licthenstien to fund and launch a manned Mars mission.

 

[Piroko]

You are assuming that what they have will sell, and that it will sell enough to keep the R&D engine running. The second assumption is far fetched.

Yea, I'm assuming that there is a large market below the Ultrabook pricepoint in which Jaguar will slot in. Richland will not be a sales wonder, but considering it's a drop-in replacement OEMs will pick it up nonetheless. Kaveri is a bit more risky, because we don't know how well it'll do against Haswell and Broadwell. The console chips should be through the R&D stage by now. The heavy lifting of GCN has been done and HD7000 is well recieved.

That should summarize most of their business. I can't tell if it's enough or not and honestly, I wouldn't bet money on anyone of you (or me) being spot on in any financial future analysis beyond the next 12 months. The last 20 years have shown me that noone had a hit rate of more than 50% in long term prediction.

 

[Fjodor2001]

Transistor cost is not the same.

According to your own diagram it is about the same.

And Qualcomm is a small company compared to Intel in revenue.

Intel 2012 Revenue: $53.3 billion
Qualcomm 2012 Revenue: $19.1 Billion

So Qualcomm is actually not that small. Also you should actually add up the revenue of all GloFo's customers to get the revenue that the R&D cost for the process tech will be spread out over.

Their [AMD's] CEO saying its done for with big cores and so on.

That is your interpretation of his statements. Others here disagree. And AMD also has a roadmap disagreeing with you.

What more do you need as edvidence?

Evidence of what? That AMD is in a tough position? None, since that we already know that. But that does not mean that they definitely will be dead in a few years time like you seem to think.

The question is what will happen going forward. They are still around despite that they should have been dead 20 years ago, to the frustration of doomsayers as yourself. This is a fast changing business. A company can go from a giant to a dwarf in a couple of year's time, and vice versa. Just look at Apple vs Nokia for example. Apple was more or less dead before Microsoft saved then financially in 1997, and now Apple is a giant. The opposite happened to Nokia. Sony was also mentioned as an example. So it's kind of naive of you to deliver such definitive predictions of what will happen to the companies in this business.

GloFo is a terrible foundry in tersm of delivery of nodes. They are already way behind and constantly changes their roadmaps to reflect it.

So you say the 6 months that AMD gets on 28nm before Intel releaes 14nm makes a difference? When will AMD hit 20nm? 2015? 2016? 20nm vs 10nm?

And people dont care if poor AMD is behind in processnodes. They will buy a 22 or 14nm product with much better power and performance charateristics. Just as they do today and why AMD is failing so hard in the CPU division.

The only question is, will AMD go belly up or end like VIA.

Well we've already concluded that moving to a later node no longer brings as many advantages as before.

We're also seeing Intel bringing ~8% CPU performance increase per year, while AMD is bringing 20-30% (although to be fair measured at a favorable period for AMD). In single threaded performance IB is only 50% faster than Trinity. If you project the current rate of CPU performance increase 2-3 years ahead, the Intel performance lead is gone.

That may perhaps not happen, but chances are that AMD will be pretty close. And then they are in a position to sell CPUs that are good enough for most people, at a competitive price. You have yourself numerous times said that the CPU performance is reaching a level where it is good enough for the average Joe, so then why should they pay more to get an Intel CPU?

Actually AMD could in a few years time be in a very good position to sell "value" desktop PCs that are good enough for most people. Just look at for example the TV business. The value and mid range segment is where all the volume is nowadays.

 

[AtenRa]

The debate was if AMD makes enough money to feed its R&D for new IC designs. The answer is, they have the R&D budget to design new Big-Core IC even at 20nm. Now, if they going to make it or not is another thing, the R&D budget is there.

Since developing and fab cost rises for each new node, Both AMD and Intel and everyone else, will have to keep using the same node for more than 2 years until they will transition to the next. Intel will start to do the same thing soon(they have already doing it at 22nm).

Intel's 22nm first product(IvyBridge) was released in January 2012, 15 months later and Haswell is still 3 months away. Intels next node will come more than 2 years later, i will not be surprised if Broadwell will not be released until early 2015(3 years after Ivy at 22nm).

 

[ShintaiDK]

According to your own diagram it is about the same.

Intel 2012 Revenue: $53.3 billion
Qualcomm 2012 Revenue: $19.1 Billion

So Qualcomm is actually not that small. Also you should actually add up the revenue of all GloFo's customers to get the revenue that the R&D cost for the process tech will be spread out over.

That is your interpretation of his statements. Others here disagree. And AMD also has a roadmap disagreeing with you.

I just consider you in denial here.

Evidence of what? That AMD is in a tough position? None, since that we already know that. But that does not mean that they definitely will be dead in a few years time like you seem to think.

Arh please, its simple math as IDC also stated.

They got 2 option. Belly up or the VIA route. And Rory tries to get them on the VIA route before its too late. Seems VIA for you is dead with that statement of yours?

The question is what will happen going forward. They are still around despite that they should have been dead 20 years ago, to the frustration of doomsayers as yourself. This is a fast changing business. A company can go from a giant to a dwarf in a couple of year's time, and vice versa. Just look at Apple vs Nokia for example. Apple was more or less dead before Microsoft saved then financially in 1997, and now Apple is a giant. The opposite happened to Nokia. Sony was also mentioned as an example. So it's kind of naive of you to deliver such definitive predictions of what will happen to the companies in this business.

Why should they have been dead 20 years ago? Can you show me they should have been dead 20 years ago? It seems to be fictional statement you create.

So you put your faith in a miracle? Great...

Well we've already concluded that moving to a later node no longer brings as many advantages as before.

Its still a disaster for the competing companies. Nomatter how you try spin it.

We're also seeing Intel bringing ~8% CPU performance increase per year, while AMD is bringing 20-30% (although to be fair measured at a favorable period for AMD). In single threaded performance IB is only 50% faster than Trinity. If you project the current rate of CPU performance increase 2-3 years ahead, the Intel performance lead is gone.

AMD didnt bring 20-30%. Your entire statement is based on bogus and false hopes. Not to mention AMD CPUs already run out of spec to even archive what they do. Cinebench single thread went from 1.02 to 1.1. Multithreaded went from 5.99 to 6.89. But thats with going from 125W to 140W+ by adding a ~11% base clock increase while turbo clocks stayed the same

That may perhaps not happen, but chances are that AMD will be pretty close. And then they are in a position to sell CPUs that are good enough for most people, at a competitive price. You have yourself numerous times said that the CPU performance is reaching a level where it is good enough for the average Joe, so then why should they pay more to get an Intel CPU?

Actually AMD could in a few years time be in a very good position to sell "value" desktop PCs that are good enough for most people. Just look at for example the TV business. The value and mid range segment is where all the volume is nowadays.

Good enough? Do I need to remind you of a -37% YoY CPU revenue?

It would be much easier if you didnt have to pass the 5 stages of grief first.

 

[Arachnotronic]

Also, AMD is fighting with 28 nm parts (Trinity) vs 22 nm parts (IB) for Intel. If you're comparing Kaveri (28 nm) vs Broadwell (14 nm) I think that is unfair, since Kaveri is expected to be released in late 2013 vs Broadwell in mid-2014. Also, you are comparing at a point where the node difference is at its peak, with Broadwell being a tick meaning that Intel will just have switched over to 14 nm vs Kaveri that likely is at the end of the 28 nm tock cycle for AMD.

Trinity is 32nm.

 

[mrmt]

That should summarize most of their business. I can't tell if it's enough or not and honestly, I wouldn't bet money on anyone of you (or me) being spot on in any financial future analysis beyond the next 12 months. The last 20 years have shown me that noone had a hit rate of more than 50% in long term prediction.

They are investing LESS than before, meaning that they should be get LESS than what they got before. If you think they can just invest less and have. Bigger. Returns, you are expecting for a miracle, and once you go to the supernatural realm, there isn't much to discuss, is it?

 

[frozentundra123456]

The discussion has become so polarized, but there is a third option, and in my mind, the most likely one. That is that AMD will continue to exist, making just enough money to survive, perhaps with help Abu Dubai. They will have to find a niche market and compete on price/performance and better igp. The only really good market I see for this is tablets, and AMD could have an advantage there for a year or two. Whether they have the marketing clout to leverage that advantage and force manufacturers to build good quality platforms that utilize their chips properly is the critical question. They apparently did not do this well with Brazos, because OEMs insisted on putting it in full size notebooks and even desktops, where there are much better alternatives.

There is also a question as to how long they can maintain the advantage they may have in the low power APU area, because Intel is attacking it from the high end by with Haswell/Broadwell and future shrinks and from the low end by improving atom.

 

[Fjodor2001]

Qualcomm Revenue 2012: $19.1 billion:
http://www.qualcomm.com/media/relea...h-quarter-and-fiscal-2012-results-fiscal-2012

Intel Revenue 2012: $53.3 billion:
http://www.intc.com/financials.cfm

I just consider you in denial here.

As you've noticed it's not just me that disagrees with you. But I guess everyone here that disagrees with you is in denial?

Arh please, its simple math as IDC also stated.

They got 2 option. Belly up or the VIA route. And Rory tries to get them on the VIA route before its too late. Seems VIA for you is dead with that statement of yours?

Others here disagree with your math, conclusions, interpretation or Rory's statement, and possible options for AMD. It's funny that in such an uncertain business, you still think that there is only one single predestined possibly outcome.

Why should they have been dead 20 years ago? Can you show me they should have been dead 20 years ago? It seems to be fictional statement you create.

Doomsayers as yourself have predicted AMD's imminent death for 20 years.

So you put your faith in a miracle? Great...

Not a miracle. It's just the nature of this fast changing business.

You would probably be the financial advisor telling their clients to sell their Apple stock in 1997, and go all in in Nokia in 2006.

Its still a disaster for the competing companies. Nomatter how you try spin it.

But no more than for Intel. That's the point.

AMD didnt bring 20-30%. Your entire statement is based on bogus and false hopes. Not to mention AMD CPUs already run out of spec to even archive what they do. Cinebench single thread went from 1.02 to 1.1. Multithreaded went from 5.99 to 6.89. But thats with going from 125W to 140W+ by adding a ~11% base clock increase while turbo clocks stayed the same

See: http://www.anandtech.com/show/6347/a...sktop-part-2/2
Single threaded performance went up 21% going from Llano (A8-3850) to Trinity (A10-5800K).

Also see: http://www.anandtech.com/show/6347/a...sktop-part-2/3
SYSMark 2012 Overall performance up 29% going from Llano (A8-3850) to Trinity (A10-5800K).

TDP was unchanged @ 100 W.

Good enough? Do I need to remind you of a -37% YoY CPU revenue?

We're talking about the future, not the past. The point is that AMD may have a chance to catch up in the coming years, since Intel is increasing the CPU performance at a lower rate than AMD currently.

Also, does your statement mean that you no longer claim the CPU performance to be good enough for most people?

Then it's funny how you change your opinion depending on how it suits your current argument. In other threads when discussing why Intel lately have produced so crappy CPU performance increase between CPU generations you say it's because the CPUs are already fast enough for most people, so they focus their efforts elsewhere. Then now when we're comparing AMD against Intel you claim the fast-enough argument to be moot. You're a funny guy. Please make up your mind...

 

[jpiniero]

That is that AMD will continue to exist, making just enough money to survive, perhaps with help Abu Dubai

Will Abu Dubai really bail out AMD? We're talking about $2B in debt.

 

[2is]

Others here disagree with your math, conclusions, interpretation or Rory's statement, and possible options for AMD.

If by "others" you mean "one other", I suppose you're correct, though "others" suggest plural and that's not the case since it seems to be you and one other person.

Most of what he has said doesn't even make sense, and most of what you're saying is based on what you want/wish to happen.

We're talking about the future, not the past. The point is that AMD may have a chance to catch up in the coming years, since Intel is increasing the CPU performance at a lower rate than AMD currently.

Lets take this for example, what exactly makes you think the future is anything but bleak for AMD? They are selling less products, making less profit, posting higher losses, investing less year after year when this market dictates you need to invest more to keep up. You can choose to be ignorant to all these factors and turn a blind eye to them, but don't expect any sensible person to follow your lead.

The enthusiast in me loves what you stand for. I would love for AMD to pull an Athlon64 out of their hat and continue to pound it out with Intel, but I won't let that feeling compromise my view of reality.

 

[vampirr]

2is... Jim Keller joined AMD, he previously worked for AMD and he designed Athlon 64 k8 and Jim Keller worked for Apple. Hes again working for AMD and he is making a revision of Steamroller design, you can officially expect another Athlon 64 even soon by end of 2013.

 

[Homeles]

Yep. Jim Keller is a superhuman that defies the laws of the universe.

/s

 

[2is]

2is... Jim Keller joined AMD, he previously worked for AMD and he designed Athlon 64 k8 and Jim Keller worked for Apple. Hes again working for AMD and he is making a revision of Steamroller design, you can officially expect another Athlon 64 even soon by end of 2013.

Unlikely. The reasons why are already in this thread.

 

[mrmt]

2is... Jim Keller joined AMD, he previously worked for AMD and he designed Athlon 64 k8 and Jim Keller worked for Apple. Hes again working for AMD and he is making a revision of Steamroller design, you can officially expect another Athlon 64 even soon by end of 2013.

Is he? Source? Or again your self-made hype?

 

[Sleepingforest]

Keller indeed was at AMD for a year, from '98 to '99, and was the lead architect on K8. He then left to join a smaller group, which was bought by Apple (and subsequently, Keller worked on A4 and A5). He rejoined AMD in 2012. However, that means whatever he's been working on is unlikely to surface for a while, since the CPU pipeline is about 5 years.

 

[frozentundra123456]

Will Abu Dubai really bail out AMD? We're talking about $2B in debt.

Who knows, but might they not at least infuse enough cash to keep AMD alive if they had a couple of bad quarters? I am not saying they would pay off their entire debt.

 

[vampirr]

Keller indeed was at AMD for a year, from '98 to '99, and was the lead architect on K8. He then left to join a smaller group, which was bought by Apple (and subsequently, Keller worked on A4 and A5). He rejoined AMD in 2012. However, that means whatever he's been working on is unlikely to surface for a while, since the CPU pipeline is about 5 years.

I am not talking about new architecture or something, Jim Keller will improve Steamroller most likely. Tweak it and maybe couple of improvements. Steamroller production will start in maybe in Q3 to meet up with demands when it gets released in Q4 of this year...

 

[Phynaz]

2is... Jim Keller joined AMD, he previously worked for AMD and he designed Athlon 64 k8 and Jim Keller worked for Apple. Hes again working for AMD and he is making a revision of Steamroller design, you can officially expect another Athlon 64 even soon by end of 2013.

Yep, because one guy can design a cpu in a year.

You know he is a manager and not a designer, right?

 

[vampirr]

Who knows, but might they not at least infuse enough cash to keep AMD alive if they had a couple of bad quarters? I am not saying they would pay off their entire debt.

Most likely scenario they will inject some serious cash to fund R&D and pay off some debts that AMD has, but if they decided to pay off the whole debt and give 1B$ then it would turn out good for everyone even thought its unlikely:ninja:

edit: Phynaz, he is a computer architect...

 

[Fjodor2001]

Lets take this for example, what exactly makes you think the future is anything but bleak for AMD? They are selling less products, making less profit, posting higher losses, investing less year after year when this market dictates you need to invest more to keep up. You can choose to be ignorant to all these factors and turn a blind eye to them, but don't expect any sensible person to follow your lead.

The enthusiast in me loves what you stand for. I would love for AMD to pull an Athlon64 out of their hat and continue to pound it out with Intel, but I won't let that feeling compromise my view of reality.

I'm not denying AMD is in a tough position. But at the same time I'm also looking at the actual performance improvements of the products that AMD and Intel bring out. And there it seems like Intel is stagnating performance wise, while AMD has been doing better.

There is still a CPU performance gap, which likely also explains AMD's poor economic figures the last couple of years. If AMD is too far behind, they simply will not sell (or only at heavily discounted prices). But if the gap narrows, then AMD may become an option again for a quite large market segment. Especially sine a lot of people think that CPUs are already getting fast enough for their use cases.

I'll also repeat what I said in an earlier post to clarify:

"It's pretty safe to assume that Intel and AMD learns from each other's designs. They cannot steal patents straight off, but can get "heavily inspired". That cannot be done without some lag however. So let's assume AMD is 2-3 years behind Intel in the CPU design area. Based on this:

a) If Intel increases CPU performance by 25% per year, that means Intel has:
1.25^2 to 1.25^3 => 156% to 195% the performance of AMD (being at 100%).

b) If Intel increases CPU performance by 8% per year, that means Intel has:
1.08^2 to 1.08^3 => 116% to 126% the performance of AMD (being at 100%)."

So if Intel is 50-100% faster than AMD, it might not be an option for most people. But if the gap narrows to 15-40% I think it's a different story. Especially when the better iGPU and lower price of the AMD CPUs is taken into account.

In other words, Intel will have to keep a substantial CPU performance improvement rate or the competition might catch up. Looking at the current trend where Intel improves CPU performance by ~8% per year, it's not completely unlikely that it will happen.

 

[ShintaiDK]

You simply cherry picked a benchmark with 2 selected CPUs. Why dont you use all the other numbers? Because they destroy your manipulating picture?

How about that?

Let me do what you do:

Haswell is now 20-75% faster than IB!

 

[Phynaz]

edit: Phynaz, he is a computer architect...

Read:

http://www.amd.com/us/press-releases/Pages/JimKellerJoinsAMD-2012aug01.aspx

VP's don't do design. "Cheif's" don't get in the trenches. They are managers.