How long can intel maintain such a large process lead over rivals?

[ShintaiDK]

It would be interesting to break down the costs of building a FAB into sub-costs. How can it cost 7 billion dollars to build a 22 nm chip factory? I mean after all it's just a building, with clean room facilities, and production lines with machines for producing chips?

Intuitively, one would expect the research costs to be much higher than the cost of producing the FAB. Once everything works, and you know how to produce those production line machines you just have to mass produce them, put them in a factory building and you're done? How can such a FAB cost 7 billion dollars?

I know things are always more complicated than they original seem, but still...

Manufactoring equipment gets increasingly more costly. And the fabs needs higher and higher standards as well.

 

[dma0991]

I know things are always more complicated than they original seem, but still...

I'm guessing equipment costs would take a huge chunk of the pie.

 

[Charles Kozierok]

If I drive you out the circuit in a car race rendering your car way slower than mine you won't keep the pace even if I get a stop and go penalty. If I do it every single lap you can guess the outcome.

I have no idea what that has to do with the current situation.

AMD had its best successes during the time when Intel was doing most of its evil stuff. Intel has been winning over the last few years by putting out better products than AMD.

It's really that simple.

 

[Imouto]

I have no idea what that has to do with the current situation.

AMD had its best successes during the time when Intel was doing most of its evil stuff. Intel has been winning over the last few years by putting out better products than AMD.

It's really that simple.

No, it's not.

As far as I know Intel has been walking the courts worldwide till late 2009 being guilty in all of them and having a better desing does nothing aganist this kind of shit.

 

[Fjodor2001]

Manufactoring equipment gets increasingly more costly. And the fabs needs higher and higher standards as well.

Assuming that is the case, it would be interesting to know why the equipment is so costly? Does it require high precision mechanics or similar? Does the equipment contain many more parts than the previous generation of manufacturing equipment? Or something else?

You can check out this video from inside the Intel 22 nm manufacturing fab:

http://www.youtube.com/watch?v=lrsPzbUJwl8&feature=related

Looks pretty high tech... But still most of it seems to consist of moving wafers and chips between different manufacturing steps (and verifying their functionality). But this was done in previous manufacturing fabs too. So I still don't see how the equipment can get THAT much more expensive with each generation...?

 

[Charles Kozierok]

As far as I know Intel has been walking the courts worldwide till late 2009 being guilty in all of them and having a better desing does nothing aganist this kind of shit.

Funny, because I've researched this pretty extensively and never found anything like that.

I won't defend some of Intel's past actions, but none of it changes the simple reality that since 2005 or so they have led technically and deserve to be in the position they are in now.

 

[blastingcap]

As long as capitalism lives, companies like Intel will be on top. As long as engineers seek only the jobs with the highest pay, Intel will win. As long as we are robots, Intel will win.

ASML comes to mind as a classic example of why this kind of thinking is wrong. (A less positive example: Kodak)

 

[ShintaiDK]

Assuming that is the case, it would be interesting to know why the equipment is so costly? Does it require high precision mechanics or similar? Does the equipment contain many more parts than the previous generation of manufacturing equipment? Or something else?

You can check out this video from inside the Intel 22 nm manufacturing fab:

http://www.youtube.com/watch?v=lrsPzbUJwl8&feature=related

Looks pretty high tech... But still most of it seems to consist of moving wafers and chips between different manufacturing steps (and verifying their functionality). But this was done in previous manufacturing fabs too. So I still don't see how the equipment can get THAT much more expensive with each generation...?

The litho tools gets more costly, not to mention more steps in the processline. Nikon/ASML supplies the litho tools for example. And alot of it needs to be modified etc. It all adds up to cost.

Same with the purified water needing to wash chemicals of the wafers for each litho step. Everything cost, cost and then cost some more. Cleaner, more pure etc. It all adds up and you cant really point to a single thing and say this is what adds the cost.

 

[pelov]

Some other interesting reads.

Fully-depleted wafers from Soitec, pre-integrate critical characteristics of the transistor within the wafer structure itself. Soitec’s FD wafers offer an early, low-risk migration at the 28nm node down to 10nm and beyond, lowering costs and enabling significant advances in the performance and power efficiency of mobile devices such as smartphones and tablets.

http://www.xbitlabs.com/news/other/..._Fully_Depleted_SOI_with_14nm_20nm_Chips.html

 

[soccerballtux]

I think Intel is hitting a wall because of the heat density.

if that were the case they wouldn't have switched to a crappy thermal compound instead of solder

 

[Imouto]

Funny, because I've researched this pretty extensively and never found anything like that.

LOL, wikipedia is too deep for you? Take it easy, here's a link:

http://en.wikipedia.org/wiki/Intel#Anti-competitive_allegations

As for sources you can check the bottom.

I won't defend some of Intel's past actions, but none of it changes the simple reality that since 2005 or so they have led technically and deserve to be in the position they are in now.

Investing and having no profits because Intel is bribing every single OEM to not mount any of your chips lead to no money to spend in research, design and hiring.

So again, no, they don't deserve their current position.

 

[ShintaiDK]

Blablabla

Cant you go derail some other thread with your crap?

 

[Arachnotronic]

Since the future in this field will be graphene based I think that Samsung will have a huge lead. They have way more money to burn in development as they're doing atm researching graphene.

Intel and IBM are researching aswell, but Samsung is really interested in graphene since they can make touchscreens, NAND, batteries and a lot of stuff out of it.

I'm sorry, what? Samsung's semiconductor manufacturing process *today* is nowhere near Intel's. And it's silly to think that Intel doesn't have a whole bunch of research projects and ideas brewing. Intel also has a lot more at stake in the high end semiconductor manufacturing biz, so they're going to spend a lot more than Samsung on that front, because it makes financial sense to do so.

I would be very surprised if Intel lost its semiconductor technology lead.

 

[ShintaiDK]

Lets summon up. Glofo ships 32nm. Samsung ships 32nm, TSMC ships 28nm, Intel ships 22nm.

When did the above companies ship HKMG? Trigates?

Intel is around 3½ years in front of TSMC if you count all technologies. Samsung and Glofo is just hopeless behind with 4+ years behind.

 

[KompuKare]

So many defenders of Intel here. They may be best now but consider this: from 2003 until 2006 AMD had the superior design because NetBurst was awful and Intel were also in the Rambus scam.

Of course, Intel had the deeper pockets and hence had better fabs so were able to mitigate some of NetBurtst's awfulness by being a node ahead of AMD and also of course AMD made mistakes. But when big Intel didn’t just spend its money on fabs and R&D; they also used it to bribe manufacturers to not use AMD chips and similar stuff for which they eventually had to settle but long after the damage was done. We will never know what might have happened if for instance Dell had used AMD processors back in the P4 days. It is possible that AMD could not have coped volume-wise but it is equally possible that the extra revenue might have enabled them to compete and today we’d be looking at a 60/40 split for x86 revenue vs the current 80/20. And for the die hard Intel fans: if Intel ‘only’ had 60% marketshare do you think they could afford to cripple / segment their chips to the same extend as they presently do?

And going back further: x86 is the most successful pile-of-c**p architecture of all time. Let me clarify that: 8086 was junk, 80286 was junk, 80386 was junk, 80486 was junk, the original Pentium was junk. 'Junk' in each case being a relative term in comparison to what was available at the time: 8086/286/386 vs 68000/68020/68030 etc., Pentium vs PPC and that's without bringing any other ISAs aside from Motorola and PowerPC into the argument. Has everyone forgotten just how awful Intel's 8/16/32 bit modes were? 64KB pagefiles, shortage of registers and so on.

But because IBM choose x86 for the PC, Intel has been able to use that huge revenue to eventually out-design everyone else: Motorola, Alpha, PPC all left the PC CPU business not because their ISA couldn’t be made to compete but because they lacked the scale. Yes, Intel has some bright people but anyone who knows the history of PCs in the 80s and 90s knows that the badly designed x86 ISA has a lot to answer for.

 

[grimpr]

In short: massive volume of chips + deep pockets + the ability to hire and retain some of the brightest people in the field. They're so far ahead that they make money hand over fist, which they invest in making sure they stay so far ahead.

They are quite frankly a fab company that happens to design chips on the side so that they have something to run through their fabs.

:thumbsup:

 

[BD231]

Even if there was a company that could take over Intel in the performance segment neither you nor either company would stand to gain anything. If say they were neck and neck you'd just lend towards weakening them both by having them fight over the same market share which would inevitably bring us back to the days of rushed products like we see in the video card market.

Intel is ahead because they have so much money to blow, if they didn't they'd be in the same position as everybody else. All you can really hope for are some valiant engineers who come up with a design of there own. Even then you still have a slew of major gaps to fill. For instance it's impossible to convince most any investor foundries are a good investment, reason being, they're NOT, hence the complete lack of start ups. Another issue, nobody has the foundry tech Intel does. TSMC/Global silicon tech sucks in comparison and can only hope to mimic the great features and thermal capabilities.

AMD did 100 mph into a wall with BullDozer because the process tech they need for their complex design isn't there, nor are there enough competent engineers to do anything but build off of old tech in slap happy configurations. As of right now the only way to compete with Intel is to gain access to their fabs.

 

[Imouto]

Cant you go derail some other thread with your crap?

Nice manners.

Play nice you two
-ViRGE

 

[infoiltrator]

Precision always costs, leading state of the art precision always costs the most, and requires the most refinement. Increased transitor counts do not scale linearly either. Nor greater use of rare materials.
Corporate politics tend to be "what I can get away with".
Fact is, effective marketing also plays a big part.
Buy what you like and trust, everyone else will.
Until there is a new technology easier for someone else to impliment,

 

[blastingcap]

But because IBM choose x86 for the PC, Intel has been able to use that huge revenue to eventually out-design everyone else.

This.

 

[Charles Kozierok]

And going back further: x86 is the most successful pile-of-c**p architecture of all time. Let me clarify that: 8086 was junk, 80286 was junk, 80386 was junk, 80486 was junk, the original Pentium was junk. 'Junk' in each case being a relative term in comparison to what was available at the time...

Yes. So what?

Welcome to the business world. It's not always what is the best technical solution, it is what you do with your technologies.

x86 may not be clean or elegant, but it gets the job done.

As for the anti-trust issues, as I said, I won't defend Intel on that score. But I think people use that as an excuse to avoid dealing with the harsh reality that Intel is a very rich and successful company regardless of some of its shady moves, and it has done pretty much all the right things over the last six years to put itself in the position it is in right now.

 

[Fjodor2001]

The litho tools gets more costly, not to mention more steps in the processline. Nikon/ASML supplies the litho tools for example. And alot of it needs to be modified etc. It all adds up to cost.

Same with the purified water needing to wash chemicals of the wafers for each litho step. Everything cost, cost and then cost some more. Cleaner, more pure etc. It all adds up and you cant really point to a single thing and say this is what adds the cost.

Sure, technology continuously gets more advanced. Just look at how cameras, TVs, CPUs, and so on gets better and more advanced each year. However that does not necessarily translate into higher cost. Instead you just raise the bar for what is considered high-end vs basic tech.

So I don't see why chip manufacturing fabs should be an exception to that rule?

 

[ViRGE]

Sure, technology continuously gets more advanced. Just look at how cameras, TVs, CPUs, and so on gets better and more advanced each year. However that does not necessarily translate into higher cost. Instead you just raise the bar for what is considered high-end vs basic tech.

So I don't see why chip manufacturing fabs should be an exception to that rule?

The rise in production costs is almost entirely in fixed costs for fab equipment/setup rather than in variable costs to produce a single chip. So end products have so far absorbed the cost through higher volume and component integration.

 

[ShintaiDK]

Sure, technology continuously gets more advanced. Just look at how cameras, TVs, CPUs, and so on gets better and more advanced each year. However that does not necessarily translate into higher cost. Instead you just raise the bar for what is considered high-end vs basic tech.

So I don't see why chip manufacturing fabs should be an exception to that rule?

I think you miss an important part. CPUs have never really been so cheap as they are today. You can say the same as those you mention. Points is, its about volume. Higher volume pays for those more expensive factories.

Factories cost more and more, but if you can keep volume up you also get more money back to pay for it. Its the same reason we go from 200mm to 300mm and soon 450mm wafers.

Its just the factories are too high cost for those without the volume or profit ration to chip that is needed.

Your LCD/LED/Plasma TV would also cost alot more if it wasnt the volume as it is. Plus you see the same progress there. Just look at the japanese TV makers. They are on respiratory help. Same thign, factories are cost runs upwards. The smaller players die out because they cant afford to keep up.

Or why we only got something like 2-3 HD makers left. Or the GPU market. Or memory market with constant dying players.

There just aint room for more when you need higher and higher volume to pay for each process.

 

[Arzachel]

There are three scenarios that would lead to Intel losing their lead and the x86 market becoming far more competitive:

1) The government recognizes that Intel is too large for other companies to reasonably compete and splits Intel into several smaller companies. (Not going to happen, the best government that money can buy won't to anything to compromise their "sponsors")

2) A breakthrough in tech happens that marginalizes a part of Intel's previous investments and a company with enough funds beats Intel at adopting it. (Graphene adoption could cause something like this but on a smaller scale)

3) Intel missteps horribly and and a company with enough funds capitalizes on this.

1. just isn't going to happen, 2. and 3. can be stopped by paying off everyone that even thinks of using competing products. Just like last time.