Intel profit sinks 27% on dreadful PC sales

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Haserath

Senior member
Sep 12, 2010
793
1
81
Because if it backfires (ie: sales continue to slide) it could get ugly.

The bigger Capex is there to prevent that. If they don't spend, they'd be worse off down the road. Time is one of the most valuable assets while improving technology.

Spend less=have less competitive products, and you become AMD, except with better management.
 

SiliconWars

Platinum Member
Dec 29, 2012
2,346
0
0
The bigger Capex is there to prevent that. If they don't spend, they'd be worse off down the road. Time is one of the most valuable assets while improving technology.

Spend less=have less competitive products, and you become AMD, except with better management.

Funny how AMD manages to remain fairly competitive at certain levels even though intel outspends them by many billions every year. Ditto vs Nvidia. Makes you wonder just how far ahead they'd be had intel not had the $5 billion+ they used to bribe Dell and others.

Intel is getting ready to self destruct. They were too late to tablets and too late to phones and their entire business philosophy is a decade out of date.
 
Jan 8, 2013
59
0
0
Yeah I'm talking about the older stuff, I'm not so silly or daft as to be thinking about more recent nodes in my post above on the topic.

But the nodes that Intel has basically retired, cycled out of production because they aren't a foundry and they don't have reason to keep 130nm in production for example, it is a shame that all that process technology know-how becomes a lost accomplishment from the perspective of mankind's heritage and written record.

Its just a shame is all, and it really need not be that way. At least with NASA for example they sold off the obsolete space shuttles (yes I know this is public vs private entities) to museums for posterity sake rather than crating them up and burying them in Warehouse 13 next to the Ark of the Covenant ;)

Money rules. D:

You don't like it? show a better alternative :cool:
 

NTMBK

Lifer
Nov 14, 2011
10,522
6,039
136
I don't see anything intrinsically flawed in your reasoning, what you outlay is within the spectrum of the possible. Without knowing Intel's specifics, though, we can't really speak to whether or not this scenario is probable.

What I do suspect is that the part I bolded is a self-limiting situation because if that were to be the case (what you wrote in bold) then Intel would surely resolve that imbalance by adjusting prices for the "in-demand" ULV chips accordingly.

People can't buy what isn't available, and Intel kinda has that market to themselves for the near future owing to the very ability they have in binning such ULV products. So Intel has the ability to avoid cannibalizing sales through price control.

But what do they do to spark demand for their standard voltage SKUs? Presumably there is a reason desktop clockspeeds and performance for IB were held to comparable levels as SB while TDP was lowered.

They left themselves room to raise the perceived performance proposition with Haswell as an upgrade candidate by bumping up stock clockspeeds (and TDP) as needed to fuel a wave of consumer and corporate upgrades if need be.

Intel is in the enviable position of not being in dire straits, this affords them the luxury of not needing to micro-manage the supply picture at a high frequency with the associated risks of over-managing the chain and throwing things into wildly off-balanced oscillations (AMD serves as the best counter-example of this, ironically enough, in every way).

If actions need be taken then I have every confidence Intel will take them, but I also have every confidence that Intel isn't going to panic and run around like a chicken with its head cut off just because the global outlook takes a dip for one fiscal year. They have a far more responsible and steady hand on the rudder of S.S. Intel IMO.

Of course this leaves them vulnerable to the inevitable "black swan" business environment, which will be the iceberg that sinks the unsinkable S.S. Intel, but I don't see the waning desktop PC being that black swan. Mostly because the waning desktop PC segment is so readily visible and obvious. Usually what gets in under the radar of a big business like Intel, such that it is their own undoing, is something completely out from left field. Course you can't help the unhelpable either, Eastman Kodak ;)

Yes, they can limit the sales of their ULV parts by pushing the prices way up on them. But the issue with that strategy is that they aren't just pushing people from thin Intel laptops to thick Intel laptops- they're pushing them into ARM powered tablets. The biggest growth sector in the consumer market is for thin, low powered, affordable, long battery life devices. If Intel want to grow the x86 market and make serious headway in the tablet market, they need a low power part with good performance- that they can also build cheaply, in volume. Ivy Bridge and Haswell ULV tick the first box, and Atom ticks the second box, but neither does both. The 32nm Atom is a good stopgap measure, but that is all it is- its GPU performance is already woefully behind ARM tablets, and the CPU performance of ARM tablets is catching up fast.

It worries me that in all of their recent talks, Intel is talking up the ULV Haswells and Ivy Bridges as the silver bullet to solve the tablet market. Why are they not pushing 22nm Atom like crazy? They'll have a totally new design on a massively advanced node, which is purpose built for tablet level TDPs, on a far smaller die than Haswell. If the market is pivoting away from traditional laptop form factors towards thinner, longer lasting devices, then Intel can't just keep churning out standard voltage parts in order to cull the "golden dies" for their ULV parts.
 

NTMBK

Lifer
Nov 14, 2011
10,522
6,039
136
Actually, starting with Haswell, the 15W ULT parts are becoming mainstream chips. That's what they meant when they said chip TDPs are being focused in the "less than half the 35W" segment. 22nm enables better frequency at lower voltages, and Haswell is the chip designed for that process. 35W dual core parts are also the latest to arrive for Haswell, signaling that the focus has changed.

In Broadwell, its rumored that 35W dual core parts cease to exist, because it'll be redundant with 15W parts that can do just as well and have cTDP.

Conclusion: Haswell and Beyond = ULT parts = mainstream

This really doesn't make sense to me. :confused:

Haswell is produced on the same process node as Ivy Bridge, with a larger and more complicated core. This is a core designed to perform excellently in 130W TDP server chips, being aggressively clocked down and heavily binned in order to reach a tablet TDP. In addition there is again a massively bulked up GPU part. Nothing in that says to me that they will suddenly be able to get Haswell to consistently work at lower voltages than Ivy Bridge (i.e. improve their standard voltage to ULV ratio). Yes, there are the improvements in sleep states and moving the VRMs and PCH onto the package to help with total platform power draw, but neither of those help with the thermal limits of the core itself.

As for "15W parts that can do just as well and have cTDP"- that's simply nonsense. Those 15W parts may perform as well as a Sandy 35W part, but with the same core design on the same process node the 35W TDP part will always outperform the 15W TDP part.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
It worries me that in all of their recent talks, Intel is talking up the ULV Haswells and Ivy Bridges as the silver bullet to solve the tablet market. Why are they not pushing 22nm Atom like crazy? They'll have a totally new design on a massively advanced node, which is purpose built for tablet level TDPs, on a far smaller die than Haswell. If the market is pivoting away from traditional laptop form factors towards thinner, longer lasting devices, then Intel can't just keep churning out standard voltage parts in order to cull the "golden dies" for their ULV parts.

Well you have to remember these are things that have been planned for nearly 4 yrs now, there really isn't much room in the 11th hour to shift around product lineups and plans.

I think Intel long ago internally decided to segment Atom vs Haswell at the tablet segment, and set themselves on a course to have Haswell be their premier tablet sku whereas atom would be focused on the handset market.

Fast forward 3yrs from that internal decision and you have a slightly different reality in the tablet market, one which might make more sense now to have a premier 22nm atom SKU instead of a 22nm Haswell SKU, but Intel has spent the past 3 yrs preparing to go to battle with a Haswell SKU and they can't just change horses mid-race.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
Because if it backfires (ie: sales continue to slide) it could get ugly.

In the semiconductor industry, not having available capacity when you need it is far worse than having capacity that is underutilized.

Not having capacity is what has kept TSMC from having an extra billion per year in revenue from the Apple account. Not having capacity is what kept AMD from an extra billion (or three) in annual revenue because large OEMs could not get the supply they needed. Not having capacity is what kept AMD from getting into Apple's laptops. Not having capacity is what kept Qualcomm from defecting from TSMC to GloFo. 40nm capacity limits at TSMC is what kept AMD from capitalizing on their lead over Nvidia at the time.

My take on Intel's capacity plans is that they are obviously expecting to be the beneficiary of a windfall of orders that are coming their way (negotiations for 14nm chips with the big OEMs would be underway right now) and chance favors the prepared.

If they are not prepared to be the beneficiary of huge demand for their 14nm products then they will be too far behind the ball to ever catch up to it. But clearly they know something we don't, namely how good their 14nm is looking and how well the contractual negotiations are going with handset makers and so on.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
Did some looking at the Intel numbers, this might be more of a thought provoker.

-7% in all CPU types, including mobile; the one exception was server (Xeon)

Server chips was +4%

Overall -6%

Possibly the Cloud effect...

Companies keep buying servers, consumers dont spend. I also demonstrated this with other sectors. Private people simply keep their money in their pocket and saves up.

Danish example:
stort-fald-i-danskernes-opsparing-fig1.jpg


This is the average cash savings. Last year the savings increased with a total of 38B DKK. Or 6.78B$. Essentially >90% of that money should have been spend on consumption. Yet it was saved up.

And if we look at the US. 66 million americans are now living in poverty. I dont think they put a high priority in upgrading their PC, even if they want a new one.
 
Last edited:

mrmt

Diamond Member
Aug 18, 2012
3,974
0
76
I guess that you re so pro intel that you dont see
what is written right under your eyes..

Here it is :



Write off in older generation inventory...

Is that how you call non qualified HSW chips....??..:rolleyes:

Older generation was inserted by xbitlabs, it was not in the original transcript.
 

Acanthus

Lifer
Aug 28, 2001
19,915
2
76
ostif.org
Microsoft has a role in this slump besides the fact that most mainstream consumers are content with the capabilities of a tablet. Had Windows 8 been made properly, the sales of PCs would've at least stayed the same.

This is true, I told a LOT of people to wait for windows 8 last year. Now it rolls around and it is a pile of crap. Some of the people who waited built new windows 7 machines. Others opted to keep what they have a little longer.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,787
136
This really doesn't make sense to me. :confused:

It should. Everyone forget their 22nm announcement presentation back in May of 2011? Where they showed big gains in low voltages but smaller gains in high?

No, you are right, the 35W will have gains, but it'll be smaller than ever. Just like there's almost no reason to use desktop parts rather than mobile extreme parts that perform so close but use 30W less. With 22nm it'll be much easier for them to make ULT parts.

Price difference there is too, but it'll be much less than Ivy Bridge(I'm not talking about list pricing), because greater volume will be there. Also even on a same process, Ivy Bridge is a straight up porting to 22nm, while I assume Haswell will change it up a bit to be better fit for it.
 

CHADBOGA

Platinum Member
Mar 31, 2009
2,135
833
136
Funny how AMD manages to remain fairly competitive at certain levels even though intel outspends them by many billions every year.
You seem to forget that Publicly Listed companies are meant to make profits, not win over people who post on forums.

On that criteria, AMD is miles behind Intel.

Makes you wonder just how far ahead they'd be had intel not had the $5 billion+ they used to bribe Dell and others.
Rebates aren't bribes and with AMD being capacity constrained at the time, it would have made no difference anyway.

Intel is getting ready to self destruct.

LOL Utter nonsense.

They were too late to tablets and too late to phones

There are no significant barriers to prevent Intel from eventually dominating in these devices. Getting there first isn't a guaranteed lock in forever more.
 

SiliconWars

Platinum Member
Dec 29, 2012
2,346
0
0
I keep hearing why there are no barriers for intel, but never hear about the advantages for intel. This is the important difference.
 

2is

Diamond Member
Apr 8, 2012
4,281
131
106
I keep hearing why there are no barriers for intel, but never hear about the advantages for intel. This is the important difference.

I keep seeing your posts, but never see you say much of anything.
 

pablo87

Senior member
Nov 5, 2012
374
0
0
In the semiconductor industry, not having available capacity when you need it is far worse than having capacity that is underutilized.

Not having capacity is what has kept TSMC from having an extra billion per year in revenue from the Apple account. Not having capacity is what kept AMD from an extra billion (or three) in annual revenue because large OEMs could not get the supply they needed. Not having capacity is what kept AMD from getting into Apple's laptops. Not having capacity is what kept Qualcomm from defecting from TSMC to GloFo. 40nm capacity limits at TSMC is what kept AMD from capitalizing on their lead over Nvidia at the time.

My take on Intel's capacity plans is that they are obviously expecting to be the beneficiary of a windfall of orders that are coming their way (negotiations for 14nm chips with the big OEMs would be underway right now) and chance favors the prepared.

If they are not prepared to be the beneficiary of huge demand for their 14nm products then they will be too far behind the ball to ever catch up to it. But clearly they know something we don't, namely how good their 14nm is looking and how well the contractual negotiations are going with handset makers and so on.

Andy Grove - " capacity is strategy " circa 2002, Fortune Magazine
 

pablo87

Senior member
Nov 5, 2012
374
0
0
Companies keep buying servers, consumers dont spend. I also demonstrated this with other sectors. Private people simply keep their money in their pocket and saves up.

Danish example:
stort-fald-i-danskernes-opsparing-fig1.jpg


This is the average cash savings. Last year the savings increased with a total of 38B DKK. Or 6.78B$. Essentially >90% of that money should have been spend on consumption. Yet it was saved up.

And if we look at the US. 66 million americans are now living in poverty. I dont think they put a high priority in upgrading their PC, even if they want a new one.

Good post. You want to see a good example go to gas stations in lower income neighborhoods. everybody puts in $10 to $20. How can they afford a new PC?
 

pablo87

Senior member
Nov 5, 2012
374
0
0
Sheesh $13 Billion dollar Capex... Time to invest in Intel. They're gearing up for something.

INTEL has been gearing up for 4 years. intel is going to be negative cash flow, BEFORE buy-backs, in 2013, they might need to borrow again. PPE might finish 2013 at $33B, almost 2x what it was 4 years ago. It's shocking really when you think about it.

Hopefully there is some huge business coming in 2013/2014. I'm a little skeptical though because I've seen that giddy? exuberance from Intel a few times before.
 
Aug 11, 2008
10,451
642
126
Good post. You want to see a good example go to gas stations in lower income neighborhoods. everybody puts in $10 to $20. How can they afford a new PC?

Not to sound unsympathetic, but I bet most of them have smartphones with 100 dollar plus per month data plans. It is a matter of priorities, which is maybe a bad sign for the consumer PC market. But we still need servers to keep all the data farms going.
 

pablo87

Senior member
Nov 5, 2012
374
0
0
Not to sound unsympathetic, but I bet most of them have smartphones with 100 dollar plus per month data plans. It is a matter of priorities, which is maybe a bad sign for the consumer PC market. But we still need servers to keep all the data farms going.

Actually we only wish. I can think of worse places they're putting the money they obtain...
 

dud

Diamond Member
Feb 18, 2001
7,635
73
91
In further news:

http://news.yahoo.com/intel-may-little-choice-big-manufacturing-bet-212103968--finance.html



"Intel may have little choice in big manufacturing bet

Intel Corp's decision to spend $13 billion in 2013 to develop and build future manufacturing technology has not gone down well on Wall Street but it may be necessary if it wants to stay on top of rivals in coming years.

The top chipmaker's shares slumped nearly 7 percent on Friday, a day after executives said the company would increase 2013 capital spending from an already dizzying $11 billion.

Some analysts decried the move, saying adding new capacity should be far from Intel's mind in a waning personal computer market. Increased spending may further pressure margins and leave Intel with even more idle capacity if PC sales keep falling.

But others believe that Intel's top priority must be maintaining its technological edge, a costly but necessary endeavor that may even pay off in the long run with market share gains. Moving up the technology ladder can also deliver cost savings, helping safeguarding Intel's margins as it tries to catch up to rivals in smartphones and tablets.

"That's the bet they're making and they're all in," said Sanford Bernstein analyst Stacy Rasgon. "If you stop, TSMC and Samsung close the gap - and you're toast."

Of Intel's $13 billion capex this year, $2 billion will go toward expanding a fabrication plant, or fab, in Oregon where engineers will work on a long-term plan to manufacture microchips on silicon wafers measuring 450 mm - about the size of a large pizza.

The other $11 billion goes toward more immediate improvements in Intel's manufacturing technology, letting it build chips over the next two or three years with features measuring just 14 nanometers, and then 10 nm. The narrower the features, the more transistors can fit on a single chip, improving performance.

The newest fabs currently use 300 mm wafers, about the size of a vinyl record. Moving up in size will make room for more than twice as many chips to be etched on each, leading to cost savings.

Lowering costs will be a serious priority for Intel as it ventures into the tablet and phone markets, where chips sell for much less than in the PC industry. Intel, which has yet to make meaningful progress in mobile, stresses that its most advanced fabs have the lowest cost per chip produced.

"One of reasons why Intel is so aggressive on capital spending is to maximize the chances it has of protecting its gross margins as it moves into smaller and lower priced CPUs," Longbow Research analyst JoAnne Feeney said.

SPENDING SPREE

Intel is not the first tech company to worry Wall Street with aggressive long-term investments whose payoffs are difficult to estimate.

Investors in the past have criticized Amazon.com Inc for splurging on costly warehouses and other shipping facilities, investments that eventually paid off and contributed to rich stock valuations.

While the size of Intel's capex increase alarmed investors, the chipmaker since 2011 has been spending heavily. Intel normally pours 12 to 16 percent of its revenue into capex, but spending has been closer to 20 percent in the past two years and will probably be higher this year, Feeney estimated.

The costs of developing the new technology to use 450 mm fabs are so high that just a few companies, such as Intel, Samsung Electronics and Taiwan's TSMC, are expected to have the scale to make the jump worthwhile. Building 450 mm plants from the ground up is expected to cost $10 billion or more.

It's not just a matter of creating bigger silicon wafers. Most of the high-tech equipment - sold by the likes of Applied Materials - used in chip manufacturing has to be redesigned as well.

The transition from 300 mm to 450 mm is so expensive and complicated that the world's biggest chipmakers and tool makers are collaborating to establish new standards and timing new technology.

Intel made a $3 billion strategic equity investment last year in chip equipment supplier ASML to help fund the development of future lithography tools for 450 mm fabs, a move followed by rivals Samsung and TSMC.

Intel's Oregon plant will lead the effort to produce chips on 450 mm wafers, with other larger Intel plants upgraded in the future, Chief Financial Officer Stacy Smith told Reuters on Thursday.

Rasgon said Intel's long-term investments in manufacturing will mean more pressure on its margins over the next few years, but that its spending will help ensure it remains a major player in the chip industry over the next decade - though there's no guarantee.

"If there's any company I can look at five years from now, they'll be here and they'll be really successful at whatever they're doing. But I don't know what they'll look like," Rasgon said.

"They have to do this, but it doesn't mean I want to own the stock while they're doing it.""
 

CHADBOGA

Platinum Member
Mar 31, 2009
2,135
833
136
I keep hearing why there are no barriers for intel, but never hear about the advantages for intel. This is the important difference.
I expect in time that Intel's Atom SoC's will lead on performance/watt over its ARM rivals and too many people don't understand that if a $30 chip has a good margin in it, then it doesn't matter that it is only $30, unless it is directly replacing a more expensive chip.

Tablets and smartphones will be overall additive to Intel's revenues, because we are not going to see a 1:1 replacement of a laptop/desktop, with a tablet or smartphone.

If and when Microsoft sorts out Win 8(probably be done by releasing Win 9 in 2014) then x86 compatibility becomes a selling point.

The main reason to even want to buy a Windows tablet is for the very deep value proposition of being able to dual-purpose the tablet as a "tablet" as well as a traditional laptop (or even desktop) PC.
 

dud

Diamond Member
Feb 18, 2001
7,635
73
91
I'm glad Intel's ramped up the spending. Time to smack down the unworthy.


Without competition you will reap what you sow. A 2nd, 3rd or even 4th party involved with processor manufacturing will ensure a healthy market ... and ensure that the consumer gets the best. Either you didn't think this through or you are lacking in forethought.
 

Homeles

Platinum Member
Dec 9, 2011
2,580
0
0
Without competition you will reap what you sow. A 2nd, 3rd or even 4th party involved with processor manufacturing will ensure a healthy market ... and ensure that the consumer gets the best. Either you didn't think this through or you are lacking in forethought.
Or maybe you're just jumping to conclusions and/or wildly misinterpreting what I wrote.