cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
22,060
Intel is preparing to start construction on an expansion to its Oregon research factory known as D1X in June. D1X is Intel's most advanced research facility where technology developed there is duplicated at other Intel facilities worldwide. The expansion is likely due to Intel adopting extreme ultraviolet lithography (EUV) which requires $120 million machinery that is as big as a bus. Intel is moving ahead with the development of its 7 nm technology which requires EUV to enable the design of smaller features on chips. The existing D1X research facility takes up an area that encompasses 2.2 million square feet of space and is equivalent to 15 Costco warehouse stores. The expansion is expected to increase Intel's footprint at D1X by 1.1 million square feet. Intel could announce the expansion as early as Thursday when it releases its 2018 financial results.

Intel's D1X factory: Intel's main research factory, or fab, in chip industry parlance. D1X is where the company develops each new generation of microprocessor and begins manufacturing it. The company duplicates the production process invented in Oregon at factories in Arizona, Ireland and Israel. Dan Hutcheson, a longtime chip industry watcher and chief executive of VLSI Research, said Intel's Oregon plans don't surprise him. In semiconductors, he said, technical advancement requires significant investment. "They keep breaking these barriers," Hutcheson said. "Part of that is just making the billion-dollar bets on the future."
 
A day late and a dollar short. Doesn't matter if TSMC or Samsung's process nodes are less advanced. By the time Intel is ready to roll out Their 7nm process TSMC and Samsung will have mature 7nm processes and possibly even be producing 5nm.

I use Intel processors however, it's poetic Justice to see AMD crawl out of nowhere to gain market share and watch the once Titans of the industry struggle to remain relevant.
 
What a wasteful of $$ by Intel. They throw more money in just to catch up with AMD.

By the time their newly built fab is in business, AMD are already on 5nm in both of processor & gpu dept.

Intel have a bad habit of shooting in their boot.
 
Im SURE they are going to the same 'less advanced' process as everybody else... Really how advance can something be if it doesn't work!
 
I don't think it's wasteful at all. They are looking hard for solutions to these problems. They are looking in the wrong place, but it's not my place to interfere.
 
As long as Intel gets to EUV in a reasonable amount of time, they will be good. The nodes jumps are taking longer and there really isn't anyone doing EUV but Samsung and TSMC. Don't think there exists a company that can take on Intel's full production load. Global Foundries isn't even in the game at this level anymore. There is enough for GF at 14nm and 7nm / EUV is just really expensive.
 
As long as Intel gets to EUV in a reasonable amount of time, they will be good. The nodes jumps are taking longer and there really isn't anyone doing EUV but Samsung and TSMC. Don't think there exists a company that can take on Intel's full production load. Global Foundries isn't even in the game at this level anymore. There is enough for GF at 14nm and 7nm / EUV is just really expensive.
GF is locked at 12nm for best node. I'm talking about the time it takes to build the facility, out fit it, tape out 7nm and then work through their first engineering samples. It could take them, easily, two years to get it right before they even get to the mass production necessary for competitive volume.

They will have to rely on 14nm+++++++++ or whatever the hell they will call it, they already took a step back with their board logic from 14 to 22nm. And someday.... 10nm might be viable, lol.

They're screwed and already massively behind. Good luck to them
 
GF is locked at 12nm for best node. I'm talking about the time it takes to build the facility, out fit it, tape out 7nm and then work through their first engineering samples. It could take them, easily, two years to get it right before they even get to the mass production necessary for competitive volume.

They will have to rely on 14nm+++++++++ or whatever the hell they will call it, they already took a step back with their board logic from 14 to 22nm. And someday.... 10nm might be viable, lol.

They're screwed and already massively behind. Good luck to them

Intel is starting 10nm now. Intel's 10nm compares well to the 10nm currently being produces by it's rivals. One process node behind for a couple years isn't going to suddenly kill them. Intel can outsource to 7nm suppliers on selected products as well. They are not as bad off as you believe. Certainly not on the cutting edge as they used to be, but still in the game.
 
Intel is starting 10nm now. Intel's 10nm compares well to the 10nm currently being produces by it's rivals. One process node behind for a couple years isn't going to suddenly kill them. Intel can outsource to 7nm suppliers on selected products as well. They are not as bad off as you believe. Certainly not on the cutting edge as they used to be, but still in the game.
Ok, they have more money than God to throw at their problems. So, you're right. 20 years of near market Monopoly and zero innovation has made them ultra rich. Sure they can offload to other foundaries and I'm fairly certain they pushed some of their 14nm production to TSMC iirc. However, they're unlikely to do so for their main bread and butter products. If they do that, it's an admission there's something wrong in Intel land. Historically, they Fab their own processors. Even Micron pulled out of the optane partnership. Many other clients have bailed as 10nm took too long to be viable.

Intels 10nm process is a far cry from how advanced it was originally supposed to be. It's missed every release target and remains to be seen if it's even viable for mass production. They're jumping on 7nm because 10nm is a failure, regardless if they, finally, get it off the ground or not.

They're screwed in the short term. Not the long term. Supposedly they will be back up to volume production by the end of 2019.... With existing process techs. I expect they will recover, but the ground they lose this year may very well Mark the beginning of their decline . Especially if their graphics cards suck ass on launch.
 
  • Like
Reactions: Rahh
like this
Intels 10nm process is a far cry from how advanced it was originally supposed to be. It's missed every release target and remains to be seen if it's even viable for mass production. They're jumping on 7nm because 10nm is a failure, regardless if they, finally, get it off the ground or not.

Intel is not skipping 10nm. 10 will be fine until the 7nm plant up and running.

They're screwed in the short term. Not the long term. Supposedly they will be back up to volume production by the end of 2019.... With existing process techs. I expect they will recover, but the ground they lose this year may very well Mark the beginning of their decline . Especially if their graphics cards suck ass on launch.

Your forecast is dubious at best. AMD has a golden window of opportunity for about a year or two here. Then it shuts. Graphics is just another Intel foray, maybe it works - maybe it doesn't. Either way, as the past has indicated - that foray as many others will not break Intel if it fails. As you say earlier - "gobs of cash."

You appear to be contradicting yourself by saying "Intel is fine in the long term" but then predict "the beginning of their decline."
Which is it?
 
So are they about even with everyone else at this point? Or are they actually behind?
 
So are they about even with everyone else at this point? Or are they actually behind?
Samsung and TSMC are in production of 7nm. Intel 10nm is starting now. Intel 10nm is approx. equivalent to the 7nm of the others.
 
Intel is not skipping 10nm. 10 will be fine until the 7nm plant up and running.



Your forecast is dubious at best. AMD has a golden window of opportunity for about a year or two here. Then it shuts. Graphics is just another Intel foray, maybe it works - maybe it doesn't. Either way, as the past has indicated - that foray as many others will not break Intel if it fails. As you say earlier - "gobs of cash."

You appear to be contradicting yourself by saying "Intel is fine in the long term" but then predict "the beginning of their decline."
Which is it?

First of all, I never said Intel was skipping 10nm, just that it's been a damn problem for them. If I recall one of the articles I read, Intel was counting on 10nm being in production something like six years ago... They are currently releasing limited quantities of 10nm chips I believe in the mobile segment and iirc they released with little to no fanfare and no significant improvement over 14nm. I might be wrong, but if it was the bees knees Intel would have been shouting it from every media outlet on the planet.

How does what I said contradict myself, just because they have financial stability on a level in which they won't disappear doesn't mean that they might not lose relevance as a chip maker to other hungrier companies. With Amazon and Google producing their own Data center chips, Intel loses relevance in the data center. With AMD finally competitive and outstripping them in production and process node advancement Intel continues to lose market share in every segment they are involved in. So, they might just go the way of IBM at some unforeseen point, where the company exists and is wealthy, yet has lost so much to fierce competition and lack of innovation.
 
First of all, I never said Intel was skipping 10nm, just that it's been a damn problem for them. If I recall one of the articles I read, Intel was counting on 10nm being in production something like six years ago... They are currently releasing limited quantities of 10nm chips I believe in the mobile segment and iirc they released with little to no fanfare and no significant improvement over 14nm. I might be wrong, but if it was the bees knees Intel would have been shouting it from every media outlet on the planet.

10nm has already show benefits for Intel with "Chip-Stacking" at CES 2019. Your comments point to Intel 10nm being worhtless, pointless or not viable. All are untrue.

How does what I said contradict myself, just because they have financial stability on a level in which they won't disappear doesn't mean that they might not lose relevance as a chip maker to other hungrier companies.

I was just quoting what was there and pointed out the contradictory statements. Thanks for clarifying I guess, but I don't follow. A long term stable company that loses relevance is not a long term stable company. You follow?

With Amazon and Google producing their own Data center chips, Intel loses relevance in the data center.

AI Chips are semi-niche so far and do not compete directly with general x64 cpu sales. It is a new market and a BONUS if you can enter that market.

With AMD finally competitive and outstripping them in production and process node advancement Intel continues to lose market share in every segment they are involved in.

Not really, IT Purchasing managers can pretty old school, stubborn, clueless.. Certification of server chips can be a long drawn out process (years) and no one wants to risk their job over the new guy AMD, which if you recall - had pulled out of the server market leaving a bad taste in everyone's mouth.

So, they might just go the way of IBM at some unforeseen point, where the company exists and is wealthy, yet has lost so much to fierce competition and lack of innovation.

Intel may go the way of IBM but being one of the only two companies holding an X86 patents - don't count on it happening within several decades.
 
10nm has already show benefits for Intel with "Chip-Stacking" at CES 2019. Your comments point to Intel 10nm being worhtless, pointless or not viable. All are untrue.

I was just quoting what was there and pointed out the contradictory statements. Thanks for clarifying I guess, but I don't follow. A long term stable company that loses relevance is not a long term stable company. You follow?

AI Chips are semi-niche so far and do not compete directly with general x64 cpu sales. It is a new market and a BONUS if you can enter that market.

Not really, IT Purchasing managers can pretty old school, stubborn, clueless.. Certification of server chips can be a long drawn out process (years) and no one wants to risk their job over the new guy AMD, which if you recall - had pulled out of the server market leaving a bad taste in everyone's mouth.

Intel may go the way of IBM but being one of the only two companies holding an X86 patents - don't count on it happening within several decades.

Chip Stacking technology isn't coming out tomorrow.

I was writing a point by point response and stopped myself.

You want to be right, no problem.

I have the energy and the ability to respond but I already see where this is going, it's derailing the thread into an argument about individual beliefs. We can do that in PM or rather, not at all.
 
Chip Stacking technology isn't coming out tomorrow.

I was writing a point by point response and stopped myself.

You want to be right, no problem.

I have the energy and the ability to respond but I already see where this is going, it's derailing the thread into an argument about individual beliefs. We can do that in PM or rather, not at all.

Everyone want to be right. It doesn't mean they are. You just haven't brought up any points that are compelling enough for me to change my mind. We're just having a discussion here, that's fine if you don't have the energy or feel the need to continue. This is a discussion board and we are having a discussion. Don't take it personally.
 
Everyone want to be right. It doesn't mean they are. You just haven't brought up any points that are compelling enough for me to change my mind. We're just having a discussion here, that's fine if you don't have the energy or feel the need to continue. This is a discussion board and we are having a discussion. Don't take it personally.
I'm not taking anything personally. However, we're actually putting on a shit show by continuing down this road. In this instance I don't give a shit about being right, I love a good argument. Just i don't see this aas a good or productive one
 
I'm not taking anything personally. However, we're actually putting on a shit show by continuing down this road. In this instance I don't give a shit about being right, I love a good argument. Just i don't see this aas a good or productive one

Huh. I thought it was a semi-relevant discussion and there were some good points.
 
just used google maps to check out the existing place. Man the plumbing alone for cooling you can see is impressive

have always wanted to tour such a place.
 
while everyone here is bickering over AMD vs. Intel, I'm just happy to see that Intel is building a plant in Oregon giving more jobs to American workers.

Amazing what happens, regardless of political party, when the people in power in a country put their workers first instead of just paying lip service to the idea and funneling cash and jobs overseas...
 
What a wasteful of $$ by Intel. They throw more money in just to catch up with AMD.

By the time their newly built fab is in business, AMD are already on 5nm in both of processor & gpu dept.

Intel have a bad habit of shooting in their boot.

Except AMD doesn't dictate when 5nm is available, that belongs to TSMC, if anything AMD better hope there isn't any delay from TSMC, otherwise they will be delay as well. Last thing AMD wants is another 20nm debacle from TSMC.
 
Except AMD doesn't dictate when 5nm is available, that belongs to TSMC, if anything AMD better hope there isn't any delay from TSMC, otherwise they will be delay as well. Last thing AMD wants is another 20nm debacle from TSMC.

AMD works VERY closely with TSMC on their chip designs, so it would be evident to AMD probably pretty far in advance if there were issues coming. There is always the possiblitlity to go to Samsung instead should TSMC 5nm be delayed significantly. At least AMD has more options than Intel. They shed Global Foundries and imagine the trouble AMD would be in if they were stuck on GF 14nm process node now?
 
AMD works VERY closely with TSMC on their chip designs, so it would be evident to AMD probably pretty far in advance if there were issues coming. There is always the possiblitlity to go to Samsung instead should TSMC 5nm be delayed significantly. At least AMD has more options than Intel. They shed Global Foundries and imagine the trouble AMD would be in if they were stuck on GF 14nm process node now?
That is certainly true, but all I was saying is AMD doesn't really dictate timeline, those are really up to TSMC and that a process node delay can have huge impact for AMD. That being said, I do not believe there is a delay at TSMC for the time being but things can change rapidly in the Semiconductor industry.

I wholeheartedly agree if AMD is still with Global Foundry, yikes!
 
That is certainly true, but all I was saying is AMD doesn't really dictate timeline, those are really up to TSMC and that a process node delay can have huge impact for AMD. That being said, I do not believe there is a delay at TSMC for the time being but things can change rapidly in the Semiconductor industry.

I wholeheartedly agree if AMD is still with Global Foundry, yikes!

Yep - 5nm TSMC on track and yes, AMD is a slave to it. Fortunately..progress is good!
 
try ENGLISH and you will be more successful.

Try juggling a family, life, career, and carpal tunnell and let me know if quality of your posts take a dip?

Also before you complain about my Grammar flaws I ran away from home and hometown when I was 16 and built a wealthy successful life of my own. Later I put myself through college by working 3 jobs at one point. I'd have never succeeded without members of my church helping me at every step and none of those people would ever speak so flippantly insulting.

So not sorry if my autocorrect flaws upset your privelaged insufferable and delicate sensibilities but I will continue to take success advice from legit people with good hearts and great minds.
 
Back
Top