The reason Intel is not producing hot chips anymore

[IntelUser2000]

I do not believe the decline will come suddenly, but its definitely something the PC guys are VERY worried about.

Look at what Brian Kraznich(Intel CEO) revealed at investor meeting this year. They plan to push 10x the volumes of Tablet chips in 2014 compared to this year, but at a loss. That indicates they are trying to get significant marketshare so they can gain in the long term.

That plan is a very risky one. I think what they are trying to do is, once they get significant marketshare, they can raise the ASPs up by making really really good parts. Those that make competing ARM chips look like AMD. I think though even now, they are underestimating ARM. A7 from Apple surely caught them offguard. And the 1.5-2x gain they were projecting in only Spring of this year reduced to 1-1.3x against other competitors, with significant deficiency in graphics.

The barrier to adoption despite having a competitive SoC with Bay Trail can be described as follows:
-More and more vendors want their own SoC, obviously for control and maximum profit
-Being a newcomer, Intel needs a huge advantage, not a middling 20-30% advantage
-Brand recognition as a SoC manufacturer, in a world where its becoming harder and harder to sell to consumers for the SoC alone

Here's what's happening in the PC world:
-Ultrabooks failed, so Intel is trying to push 2-in-1s. Again, the problem that existed for Ultrabook adoption exists for 2-in-1s. That is, right pricing it. I assume if Ultrabooks had really good ones in the $799-899 range really quick, it would have at least not been called a failure. At least Bay Trail seems to have potential there.
-Haswell adoption quite frankly, sucks. Intel and the PC guys are desperate to move IVY BRIDGE volumes. Black Friday sales had discounts of 30-40%. The availability of 2-in-1s and Ultrabooks for Haswell is actually worse than with Ivy Bridge, despite Intel's claim it'll be 3x or something.
-PC guys like Acer, Asus, and HP are focusing on alternative projects like Chromebooks, because the sales are poor on Windows 8 designs. Lot of them is due to cost. Interestingly, Asus's plans of marketshare gain in 2013 faltered because of pushing "premium" systems that really weren't worth the price(like the Taichi with piss poor battery life, Zenbook Infinity with extraordinarily high pricing).

-->Tablets delayed PC purchase-->Reduction in PC ecosystem investment, both hardware and software-->Further reduced sales of PCs-->More impact of Tablets-->further reduction in PC investment.

There's a possibility of a devastating loss, like with Blackberry and Nokia. That would happen if major players like Intel suddenly makes a panic decision and say, makes a stupid announcement about "abandoning the PC" or something. But even if they don't, it seems inevitable that calculated strategies and high investments are needed in both Tablet and PC for them to at least survive, nevermind succeed.

 

[WaitingForNehalem]

I do not believe the decline will come suddenly, but its definitely something the PC guys are VERY worried about.

Look at what Brian Kraznich(Intel CEO) revealed at investor meeting this year. They plan to push 10x the volumes of Tablet chips in 2014 compared to this year, but at a loss. That indicates they are trying to get significant marketshare so they can gain in the long term.

That plan is a very risky one. I think what they are trying to do is, once they get significant marketshare, they can raise the ASPs up by making really really good parts. Those that make competing ARM chips look like AMD. I think though even now, they are underestimating ARM. A7 from Apple surely caught them offguard. And the 1.5-2x gain they were projecting in only Spring of this year reduced to 1-1.3x against other competitors, with significant deficiency in graphics.

The barrier to adoption despite having a competitive SoC with Bay Trail can be described as follows:
-More and more vendors want their own SoC, obviously for control and maximum profit
-Being a newcomer, Intel needs a huge advantage, not a middling 20-30% advantage
-Brand recognition as a SoC manufacturer, in a world where its becoming harder and harder to sell to consumers for the SoC alone

Here's what's happening in the PC world:
-Ultrabooks failed, so Intel is trying to push 2-in-1s. Again, the problem that existed for Ultrabook adoption exists for 2-in-1s. That is, right pricing it. I assume if Ultrabooks had really good ones in the $799-899 range really quick, it would have at least not been called a failure. At least Bay Trail seems to have potential there.
-Haswell adoption quite frankly, sucks. Intel and the PC guys are desperate to move IVY BRIDGE volumes. Black Friday sales had discounts of 30-40%. The availability of 2-in-1s and Ultrabooks for Haswell is actually worse than with Ivy Bridge, despite Intel's claim it'll be 3x or something.
-PC guys like Acer, Asus, and HP are focusing on alternative projects like Chromebooks, because the sales are poor on Windows 8 designs. Lot of them is due to cost. Interestingly, Asus's plans of marketshare gain in 2013 faltered because of pushing "premium" systems that really weren't worth the price(like the Taichi with piss poor battery life, Zenbook Infinity with extraordinarily high pricing).

-->Tablets delayed PC purchase-->Reduction in PC ecosystem investment, both hardware and software-->Further reduced sales of PCs-->More impact of Tablets-->further reduction in PC investment.

There's a possibility of a devastating loss, like with Blackberry and Nokia. That would happen if major players like Intel suddenly makes a panic decision and say, makes a stupid announcement about "abandoning the PC" or something. But even if they don't, it seems inevitable that calculated strategies and high investments are needed in both Tablet and PC for them to at least survive, nevermind succeed.

Having done an internship at Intel this summer, I can confirm much of this is spot on. Instead of creating state of the art CPUs that made great technological advances and raised the performance ceiling, the new focus is on lower power and reduced BOM costs. I still don't understand why Intel wants to compete in the commodity market when a company like Mediatek can produce a product that performs adequately with much less overhead. Volume can only get you so far when your margins are so little. As much as it pains me, people could not care less about what is inside their computer. So the OEMs will understandably always go with the lowest bidder. With 100,000+ employees, competing in this market will be difficult. FYI, I'm a college student so my experience is very limited however things just don't seem to add up to me...

 

[kimmel]

Having done an internship at Intel this summer, I can confirm much of this is spot on. Instead of creating state of the art CPUs that made great technological advances and raised the performance ceiling, the new focus is on lower power and reduced BOM costs. I still don't understand why Intel wants to compete in the commodity market when a company like Mediatek can produce a product that performs adequately with much less overhead. Volume can only get you so far when your margins are so little. As much as it pains me, people could not care less about what is inside their computer. So the OEMs will understandably always go with the lowest bidder. With 100,000+ employees, competing in this market will be difficult. FYI, I'm a college student so my experience is very limited however things just don't seem to add up to me...

Lets start with some facts of the market at the moment:
Server space (High Power High Cost) is growing.
Tablet/Phone (Low power Low cost Low Performance) is growing
Mainstream PC (High Power High performance Medium/Low cost) is shrinking

What would you do if you were Intel?

What I see Intel doing: They are bifurcating, as most people in this forum hate. They are focusing on growing segments of the market. Pushing the envelope with server CPUs and they are able to charge according to the performance advantage. They are also driving down costs and power levels to compete with products at the low end to compete with phones and tablets. The consumer left low/midrange (volume) Intel chips long before Intel stopped pushing performance at the expense of power and cost. People don't want to pay for dies that would increase performance. It's pretty obvious when you step back.

Intel is finally listening to what the majority of consumers want and focusing resources there. That's why they aren't producing "hot" chips anymore. (according to your definition)

 

[WaitingForNehalem]

What would you do if you were Intel?

I'd pull an IBM and invest more into the enterprise/server market. Intel dominates this sector and has very few competitors. They also have giant margins. I'd also grow the semi-custom CPU unit and possibly open up the fabs more.

For the consumer PC market, I would go back to pushing the performance envelope. I'd have the UI/Experience R&D unit research more into usage paradigms and killer apps that were previously infeasible. Advertise to the masses about gaming PCs and grow the enthusiast market (which is already much larger than most here realize). Also I'd work closely with Microsoft and the OEMs (they started doing this with the ultrabook initiative).

Finally, I would only focus on the high-end mobile market. Build high-end phones running Ubuntu that can dock and provide a full experience. People will know the Intel means quality.

 

[24601]

I'd pull an IBM and invest more into the enterprise/server market. Intel dominates this sector and has very few competitors. They also have giant margins. I'd also grow the semi-custom CPU unit and possibly open up the fabs more.

For the consumer PC market, I would go back to pushing the performance envelope. I'd have the UI/Experience R&D unit research more into usage paradigms and killer apps that were previously infeasible. Advertise to the masses about gaming PCs and grow the enthusiast market (which is already much larger than most here realize). Also I'd work closely with Microsoft and the OEMs (they started doing this with the ultrabook initiative).

Finally, I would only focus on the high-end mobile market. Build high-end phones running Ubuntu that can dock and provide a full experience. People will know the Intel means quality.

Intel and Microsoft needed to build a 6" Windows 8.1 Phablet 4 years ago. They finally have everything lined up to release one >>right now<<.
Merrifield + REAL Windows 8.1 + LTE stack + 6" 1080p screen = Businessgasm
Businesses wouldn't even have to update their Internet Explorer 6.0 apps, lulz

At that point initiate dumping mode to drive all ARM licensees out of business while suing them Apple/Rambus style (Rambus style as in planting moles in all the other companies to purposefully patent infringe at the most basic level possible irrevocably).

^^^^^^^^^^^^^^^^^^^
What Would Steve Jobs Do.

 

[mrmt]

I'd pull an IBM and invest more into the enterprise/server market. Intel dominates this sector and has very few competitors. They also have giant margins. I'd also grow the semi-custom CPU unit and possibly open up the fabs more.

IBM is a huge software company with a small hardware business.

Intel is practically a hardware-only business.

 

[kimmel]

I'd pull an IBM and invest more into the enterprise/server market. Intel dominates this sector and has very few competitors. They also have giant margins. I'd also grow the semi-custom CPU unit and possibly open up the fabs more.

So it sounds like you would do exactly what Intel is doing in the high end almost word for word.

For the consumer PC market, I would go back to pushing the performance envelope. I'd have the UI/Experience R&D unit research more into usage paradigms and killer apps that were previously infeasible. Advertise to the masses about gaming PCs and grow the enthusiast market (which is already much larger than most here realize). Also I'd work closely with Microsoft and the OEMs (they started doing this with the ultrabook initiative).

Can't say I'd agree. Virtually no consumers make purchasing decisions on performance right now. If they did Intel would be crushing ARM vendors but that isn't even close to what is happening. I think Intel mistakenly thought that phones and tablets were separate additive markets at first as well. When in reality it's all the market basic personal compute. Checking emails and Facebook and Netflix are the most demanding things that the vast majority of people use their laptops for. Maybe if the next "facebook" required a really fast cpu then it might trigger a round of upgrades. As it stands, until that killer app comes out they are better focused on real market demands and not imaginary ones. IMO

Finally, I would only focus on the high-end mobile market. Build high-end phones running Ubuntu that can dock and provide a full experience. People will know the Intel means quality.

Don't think this is different from what is going on today just it seems their products can't get to market quick enough.

 

[gipper53]

Nobody I know is buying a traditional desktop anymore. Friends, family, coworkers...they are buying laptops or tablets. Many still have a desktop, but none have bought a new one in several years. Many have purchased laptops, tablets, ultrabooks, uber-smartphones, etc. I can't remember the last time anybody told me about their new desktop, but I've seen and chatted about plenty of laptops and mobile devices this year.

I work in architecture, and my coworkers are using top end 2011 overclocked workstations 8+ hours a day. They are desktops that would make any enthusiast salivate, so they know what it's like to use a "real" machine. But when they go home they are using laptops and tablets.

I know it's anecdotal, but I would imagine this is not uncommon.

 

[VirtualLarry]

The consumer left low/midrange (volume) Intel chips long before Intel stopped pushing performance at the expense of power and cost. People don't want to pay for dies that would increase performance. It's pretty obvious when you step back.

Intel is finally listening to what the majority of consumers want and focusing resources there. That's why they aren't producing "hot" chips anymore. (according to your definition)

I've said it before: The rise of "good enough" computing is the death of moore's law. Moore's law (as far as the performance version of it) depended on the mass-market subsidizing the R&D for the high-end chips.

 

[kimmel]

I've said it before: The rise of "good enough" computing is the death of moore's law. Moore's law (as far as the performance version of it) depended on the mass-market subsidizing the R&D for the high-end chips.

Spend the process gains in cost and power reductions vs performance increases and it's certainly not dead at the low end. Moore's law says nothing about performance increases.

 

[CodeguruX]

The decline of the PC already happened in the 90's. Laptops took over as number one in resources, price, and popularity. Oh wait, that was probably from a good piece on the decline of the PC in the early 90s, because it never happened...

 

[WaitingForNehalem]

So it sounds like you would do exactly what Intel is doing in the high end almost word for word.

I'd go further than producing top notch x86 chips. Maybe offer some smart software monitoring solutions that when paired with the hardware can show useful stats and offer mesh power management. I think they could do even more than what they are currently doing. Also, I'm not sure how custom their custom foundry is. I believe you can't add custom IP blocks on top of their x86 CPUs unless you are a gigantic company. They should make the BayTrail chips modular and customizable like AMD Jaguar. I might be wrong on this so correct me if I am.

Can't say I'd agree. Virtually no consumers make purchasing decisions on performance right now. If they did Intel would be crushing ARM vendors but that isn't even close to what is happening. I think Intel mistakenly thought that phones and tablets were separate additive markets at first as well. When in reality it's all the market basic personal compute. Checking emails and Facebook and Netflix are the most demanding things that the vast majority of people use their laptops for. Maybe if the next "facebook" required a really fast cpu then it might trigger a round of upgrades. As it stands, until that killer app comes out they are better focused on real market demands and not imaginary ones. IMO

OK, so let's say Intel shifts much of its product line to the mobile market. What happens when the phone/tablet market is over-saturated and reaches the fast enough point (some would argue it already has)? The biggest thing pushing for more performance on the mobile side is...gaming. That is why with every new SoC, the GPU has a huge increase in performance and takes more die area.

Now the full tablets offering Windows 8.1 could use more performance for a while but for Android and iOS they are plenty fast at this point. Even Apple is having trouble innovating. Their newest iPad is a little faster, thinner, and lighter. Whoopty do

 

[luckyeight]

Nobody I know is buying a traditional desktop anymore. Friends, family, coworkers...they are buying laptops or tablets. Many still have a desktop, but none have bought a new one in several years. Many have purchased laptops, tablets, ultrabooks, uber-smartphones, etc. I can't remember the last time anybody told me about their new desktop, but I've seen and chatted about plenty of laptops and mobile devices this year.

I've noticed the same. I'm one of the few people among those I know who use a desktop at home. Almost everyone uses a portable device like laptop/tablet exclusively

The desktop for home use is definitely on the decline. It won't be replaced in the workplace, but at home its heyday has passed

 

[cbn]

The desktop for home use is definitely on the decline. It won't be replaced in the workplace, but at home its heyday has passed

Looking forward to seeing some Intel hardware in this form factor:

 

[24601]

I've noticed the same. I'm one of the few people among those I know who use a desktop at home. Almost everyone uses a portable device like laptop/tablet exclusively

The desktop for home use is definitely on the decline. It won't be replaced in the workplace, but at home its heyday has passed

Two of my friends are lead programmers for game studios and they both use laptops as their primary computing device ._.

 

[MrSquished]

desktops will definitely become a significantly smaller part of computing device market share. laptops are just too good at doing what most companies need them to do, you can bring them to meetings, bring them home to work, etc... and for the consumer it's all they need for most things.

however desktops will not go away. apple is one of the smartest companies around and they just released a new Mac Pro. they wouldn't have put the money into developing a high end desktop if they didn't see a good enough market for it.



I have 7 hard drives, because i edit photos and backup everything. maybe in 100 years when ssd storage becomes small and cheaper, i won't need standard drives. but graphics cards aren't getting smaller as they get better if you want to game or do serious editing. until that happens, you are gonna need a box to put 'em in, so the desktop lives on for people that game and people that do things like video or photo editing and design work.

 

[witeken]

I still don't understand why Intel wants to compete in the commodity market when a company like Mediatek can produce a product that performs adequately with much less overhead. Volume can only get you so far when your margins are so little.

There are huge opportunities for extra income. Once there are just a few companies left, margins will increase.

 

[witeken]

I've said it before: The rise of "good enough" computing is the death of moore's law. Moore's law (as far as the performance version of it) depended on the mass-market subsidizing the R&D for the high-end chips.

What is the performance version of Moore's law? I use 2 versions of moore's law:
1) The number of transistors in a chip doubles every 2 years.
2) The number of transistors in the same area doubles every 2 years.

Both mean that transistor size should become 1.4x smaller every 2 years.

I've never seen a "performance version" of it.

At least for the next 10+ years, Moore's law won't be death (1nm should arive around 2030). As long as companies like Intel can succeed to get enough money for R&D, which I don't think will be a problem for the next years.

 

[ShintaiDK]

There are huge opportunities for extra income. Once there are just a few companies left, margins will increase.

As complexity rises, the smaller companies will naturally fall out due to R&D cost. Since the segment got an upper limit in volume.

 

[witeken]

OK, so let's say Intel shifts much of its product line to the mobile market.

Intel won't shift anything to the mobile market as far as I know. They just made a new architecture for smartphones and tablets.

What happens when the phone/tablet market is over-saturated and reaches the fast enough point (some would argue it already has)?

That market already reached that point with displays with 1280x720 resolutions. Next year, you will see smartphones with 4x that resolution.

 

[daxzy]

What happens when the phone/tablet market is over-saturated and reaches the fast enough point (some would argue it already has)?

Well the thing is, tablets and especially phones are more a status symbol. People upgrade iPhones and Androids not only because they want the latest and the greatest, but also to let everyone know they have the latest and the greatest.

For the most part, no one gives a crap what desktop you have at home.

 

[Homeles]

What happens when the phone/tablet market is over-saturated and reaches the fast enough point (some would argue it already has)?

You don't seem to understand what the "fast enough" argument is about.

Let me put it this way: does the the mobile market need performance, or battery life more?

A really clear example of "fast enough" mobile computing is the MacBook Air. CPU performance stayed flat with the Haswell upgrade, but now battery life is sitting at around 13 hours or so. It can be extended to more than power as well. You could keep performance relatively flat while significantly reducing costs.

On the tablet and smartphone side of things, "HUGS," or Hurry Up and Go to Sleep, is a very important concept in the mobile world. Claiming that there's even a "fast enough" point in the first place is simply wrong, since speed is rather critical to power efficiency in these markets.

A rather critical point of these devices anyway is to do what your laptop or desktop can do, while weighing substantially less and being more portable. Faster chips help drive tablet sales.

 

[Schmeh39]

What is the performance version of Moore's law? I use 2 versions of moore's law:
1) The number of transistors in a chip doubles every 2 years.
2) The number of transistors in the same area doubles every 2 years.

Both mean that transistor size should become 1.4x smaller every 2 years.

I've never seen a "performance version" of it.

At least for the next 10+ years, Moore's law won't be death (1nm should arive around 2030). As long as compuringanies like Intel can succeed to get enough money for R&D, which I don't think will be a problem for the next years.

#2 is the correct version of Moore's Law. Moore's Law never explicitly predicted performance. I say "explicitly", because as transistors shrink you can make chips more complex, gaining performance. It also generally allows switching speed to be scaled up.

As for the future process nodes. Intel said at IDF this year that they see a clear path to 7nm in 2017. I wouldn't hold my breath waiting for 1nm. That is going to be extremely difficult and may be impossible without a major shift or change in the manufacturing tech.

 

[Mondozei]

I think desktop will die faster than notebook. Everyone at my company uses a notebook for office suite, and the only desktop left are kiosk machines. Only reason to have a desktop now is for an expensive gaming rig or niche programs that needs the horsepower.

Will notebooks go away if tablets get powerful enough. Maybe, but for now i like doing work on a big screen and keyboard, having a good docking device for tablets might change that though.

A tablet that can run all your office software but can also dock to a full size keyboard and monitor will be the end of pcs. I think I'll name this new device tabtop, or desklette, or even notelette. Trademark it now and make trillions from qualcomm

But what you have then is a desktop PC where you can take the PC with you. A tablet is a standalone device. When it is used as a computing device but the actual input is done on a keyboard and the viewing is done on a much bigger monitor then it is a PC.

Also, most people will prefer using a fullblown monitor and keyboard to a notebook. 24 inch monitor > 15 inch screen.

When we talk about tablets we should talk about using tablets as... tablets... not as a slave device for our desktop work. What I mean by that is how do you forsee people using office on their tablets with touch? Notebooks have long been powerful enough to do it and aside from your company I have seen very few use notebooks as their primary office workstation. When on the move, sure, but not in office.

Also, most research firms point to a stabilization in the overall PC market in 2015 and small growth henceforth.

Btw, best of luck with your patenttrolling!

 

[R0H1T]

You don't seem to understand what the "fast enough" argument is about.

Let me put it this way: does the the mobile market need performance, or battery life more?

A really clear example of "fast enough" mobile computing is the MacBook Air. CPU performance stayed flat with the Haswell upgrade, but now battery life is sitting at around 13 hours or so. It can be extended to more than power as well. You could keep performance relatively flat while significantly reducing costs.

And this is where reality hits Intel cause quad/octa core cortex A7 based phones/tablets are plenty fast enough & have excellent battery life.

The other thing to consider is that with bigger resolution phones/tablets entering mass market GPU is getting more attention & this is again where ARM is advancing at a faster pace than Intel, even with a node's(or two) disadvantage, so any amount of perf/watt lead that Intel had in the desktop arena, against AMD, is at best 1.1~1.3x times the most efficient ARM SoC in the mobile market.

On the tablet and smartphone side of things, "HUGS," or Hurry Up and Go to Sleep, is a very important concept in the mobile world. Claiming that there's even a "fast enough" point in the first place is simply wrong, since speed is rather critical to power efficiency in these markets.

This is wrong as has been proven by how leaky/inefficient the A15 is vs an A7.

Now I've said this before & will say it again that the display is the single most power consuming component of a phone/tablet & so whatever Intel puts out there will at best give you one half to an hour more of battery life depending on your usage.

A rather critical point of these devices anyway is to do what your laptop or desktop can do, while weighing substantially less and being more portable. Faster chips help drive tablet sales.

Again rather subjective because cheaper tablets will still sell a heck lot more, like the Nexus or Kindle fire, unless of course we're talking about Apple, where it's the brand that's selling & not necessarily the device or its performance.