Intel may dismiss tick-tock after Haswell, no performance CPUs (LGA based) anymore?

[cbn]

What's worse, they're already taking tools out of 32nm factories for 14nm -

Pablo,

Do you have a link to that info?

At this time Intel has plans to use Fab D1X (which it is building another module on to), Fab 42 (a brand new fab shell) and Fab 24.

The first two fabs are brand new. Fab 24 has been around since 2006 (and manufacturers 65nm and 90nm)

 

[cbn]

I think its almost certain if Intel starts fabbing for others majority of their advantage will go away. Their biggest strength and the reason they have such a lead is because they are such a well executed IDM. That's why I believe if something happens that breaks that trend, you'll wonder where their advantage has went. Something like the company splitting up or even fabbing for others.

The more I think about this...the more I agree with you.

If Intel actually made more money on Fabbing, that would indicate their chip design capabilities are most likely decreasing.

But if Intel's chip design is leading edge (like their fabs) then we should see Intel's own older designs as the best candidate for the older nodes.....not other company designs.

I guess it boils down to Intel balancing design vs. fab costs.

Too much fab and not enough design and we get a situation like Netburst vs. Athlon 64. (Intel had the better process tech with 65nm, but AMD beat them with a better design despite being on 90nm.)

Now if only Intel can get atom to the bleeding edge fab? I"ll bet this is where the company is going. It seems so logical considering light weight and thin (both enhanced by process tech and SOC design/integration) are very important in that category. I'm sure the subsidy of US carriers helps encourage development also.

 

[ShintaiDK]

8 series chipsets will be made on 32nm for example.

 

[grimpr]

Charlie D. at S|A confirms too.

Intel kills off the desktop, PCs go with it
Analysis: Broadwell has no socket, PCs have no relevance

http://semiaccurate.com/2012/11/26/intel-kills-off-the-desktop-pcs-go-with-it/

 

[ShintaiDK]

Charlie D. at S|A confirms too.

LoL, Charlie steals an article and pretends its all himself.

 

[VirtualLarry]

And of course, the rumored demise of AMD has nothing to do with this change in philosophy by Intel, right?

 

[ShintaiDK]

And of course, the rumored demise of AMD has nothing to do with this change in philosophy by Intel, right?

You forgot to add that the last part is a rumour

 

[aigomorla]

man im waiting for the day the desktop is all integrated on 1 pcb down to the ram / gpu and everything.

then we will be on the forums with the "remember the old days" when you had to select ram, select gpu, select cpu, and it ran as much wattage as a space heater, generated the same amount of heat, instead of these low watt boxes.

 

[ShintaiDK]

Back to optional L2 with SRAM chips, sockets for FPUs, 3rd party sound, NIC and disk controller! All those evil companies steals our options!

/Lulz...

 

[Lepton87]

Well, I bought 2500k instead of 2600k thinking that IVY would be a minor step forward both in IPC and OC headroom. It turned out that it indeed has a healthy IPC boost for a tick but the OC headroom actually decreased barring some extreme cooling solutions and removing IHS. Even after Intel cheap move to not use solder is taken care of it still does not reach the same frequencies as sandy using AIR or even plain water. I'm mad at myself that I didn't buy 2600k right from the start.

 

[exar333]

“I, for one, welcome our new All-in-one PCB overlords”

 

[Fjodor2001]

The article seems pretty solid. It is now also updated with the following comment:

"Updated 11/26/12@3:25pm: One point to add, two OEMs have confirmed to SemiAccurate that they have now been briefed that Broadwell is BGA only. This was done weeks after we first told them about the problem."

I wonder how this will affect motherboard manufacturers?

 

[cbn]

The article seems pretty solid. It is now also updated with the following comment:

"Updated 11/26/12@3:25pm: One point to add, two OEMs have confirmed to SemiAccurate that they have now been briefed that Broadwell is BGA only. This was done weeks after we first told them about the problem."

I wonder how this will affect motherboard manufacturers?

http://forums.anandtech.com/showpost.php?p=34291877&postcount=55

 

[IntelUser2000]

And of course, the rumored demise of AMD has nothing to do with this change in philosophy by Intel, right?

The truth is that Intel did focus too much on going against AMD rather than looking more long term. Now I look at it, I don't think AMD was ever the problem. Other than the maligned Netburst days and that's mostly due to Intel's fault.

After Core 2 it should have been about using similar cores(so you'd still see Nehalem, Sandy Bridge, etc) but on a perf/watt process rather than one aimed at absolute performance(therefore frequency). It's happening at 22nm already, where Ultrabook-class chips and Atoms will see significant performance increase and power reductions. But not so much on the traditional high end because the process isn't made for absolute performance anymore.

I guess nobody saw it though really. The clue should have been back with Core 2 when people started saying we finally had "enough" performance. Then the project goals would have been how do we advance CPUs without emphasizing solely on performance anymore?

 

[pablo87]

Pablo,

Do you have a link to that info?

At this time Intel has plans to use Fab D1X (which it is building another module on to), Fab 42 (a brand new fab shell) and Fab 24.

The first two fabs are brand new. Fab 24 has been around since 2006 (and manufacturers 65nm and 90nm)

Heard it from the CFO on the Q3 conference call.

 

[Arachnotronic]

The truth is that Intel did focus too much on going against AMD rather than looking more long term. Now I look at it, I don't think AMD was ever the problem. Other than the maligned Netburst days and that's mostly due to Intel's fault.

After Core 2 it should have been about using similar cores(so you'd still see Nehalem, Sandy Bridge, etc) but on a perf/watt process rather than one aimed at absolute performance(therefore frequency). It's happening at 22nm already, where Ultrabook-class chips and Atoms will see significant performance increase and power reductions. But not so much on the traditional high end because the process isn't made for absolute performance anymore.

I guess nobody saw it though really. The clue should have been back with Core 2 when people started saying we finally had "enough" performance. Then the project goals would have been how do we advance CPUs without emphasizing solely on performance anymore?

We have seen this with Sandy Bridge/Ivy Bridge in mobile...

 

[cbn]

Heard it from the CFO on the Q3 conference call.

Thanks for the information.

http://seekingalpha.com/article/928...-2012-results-earnings-call-transcript?page=1 <---Transcript from Intel 2012 Q3 conference call.

I didn't see specific mention of 32nm equipment being replaced with 14nm. (The term "older generation technologies" was used instead, but Its not clear what node or nodes that refers to.)

Stacy Smith does mention too much 22nm inventory is a problem. (see 2nd quote). And if you look at the first part of the third quote I provided I just wonder if *some* 22nm equipment is getting replaced or converted to 14nm?

Regarding 32nm, I do know that node will still be used for Haswell and Broadwell PCH so keeping some of those fabs open will be part of Intel's plan.

Some quotes from Stacy Smith about 14nm:

As a result of weaker than expected demand environment, we are taking several actions. We significantly cut factory loadings at the end of the quarter and we will maintain low factory utilization rates throughout the fourth quarter. We expect these factory adjustments to help bring down our total inventory levels by approximately $0.5 billion. Additionally, we are redirecting equipment and space to 14-nanomter from older generation technologies. The result of this is a $1.2 billion decrease from our July forecast for capital spending with the midpoint of our capital spending forecast for the year now at $11.3 billion.

One is we are taking down utilization in the factories down to sub 50%, again to take inventory out and free up the opportunity to move both space and equipment and redirect that to 14-nanometer. So it&#8217;s a pretty significant series of actions. And I also want to point to the inventory that we have in place while it&#8217;s in terms of units more than I want to hold. It&#8217;s on the order of 70% Ivy Bridge, so it&#8217;s our freshest stuff. I am not worried about the salability of the inventory, but I do want to bring the Ivy Bridge inventory levels down. It&#8217;s just healthy for us to have less.

I would say it&#8217;s more down the line of my presentation where I showed you that we're constantly trying to match our capacity that's in place to the demand. We are putting in capacity for bigger second half than we got. We are going to now make adjustments, so that we can move capacity to 14-nanometer and bring our inventory levels down.

It doesn&#8217;t change the fact that our planning model, we&#8217;re always looking to make sure we have the ability in terms of what we call white space to some unallocated factory space as well as some equipment to respond to upsides. The risk of being caught short and the cost of being caught short is much more than the cost of being long, because as you can see when we get a little long, we can take actions and within a six month period, we can get back aligned. If you are sure it can take you two years to get caught back up.

Some interesting commentary (from the Intel 2012 Q3 conference call transcript) on re-using 22nm equipment for 14nm:

Kevin Cassidy - Stifel Nicolaus

And with the reuse of equipment, can you give an idea what percentage that is that can be reused from 22-nanometer to 14-nanometer?


Stacy Smith - SVP, CFO & Director, Corporate Strategy

It&#8217;s not different than our historical pattern. So you could think of it in the range of 80% to 90% of the equipment that we buy and 22-nanometer is usable at the 14-nanometer nodes. And that&#8217;s not by accident. It&#8217;s one of the things that technologists spend a lot of time to make sure we have these forward reuse paths because it gives us great flexibility to respond to things like we&#8217;re seeing today.

 

[Haserath]

man im waiting for the day the desktop is all integrated on 1 pcb down to the ram / gpu and everything.

then we will be on the forums with the "remember the old days" when you had to select ram, select gpu, select cpu, and it ran as much wattage as a space heater, generated the same amount of heat, instead of these low watt boxes.

It will be sad when the enthusiast sector dies out though. Not much to differentiate from if they take all of that away.

 

[DocJones]

Another source:

http://www.extremetech.com/computin...end-of-socketed-cpus-the-end-of-upgradable-pc

 

[ShintaiDK]

Another source:

http://www.extremetech.com/computin...end-of-socketed-cpus-the-end-of-upgradable-pc

Its not another source when you keep quoting the same original article.

 

[Torn Mind]

It will be sad when the enthusiast sector dies out though. Not much to differentiate from if they take all of that away.

More like "IF". Note that there is a "professional" sector that demands powerful chips like the i7s 3770, LGA 2011 i7s, or their Xeon equivalents such as big TV companies who need to quickly distribute their broadcast segments onto the Internet or perhaps some souped up CAD programs. And of course, I don't see gamers suddenly dropping off the planet either.

 

[krumme]

So what has the cost of 22nm been so far, $24B? At least that was the capex total for 2011, 2012 a two fold increase from historical. What's worse, they're already taking tools out of 32nm factories for 14nm -

It doesn't seem like it was worth the performance improvement to go to 22nm, putting IB on 32nm might have worked too.

In any event, Intel is spending too much capex money, they need to change the approach because the status quo doesnt' work bottom line wise. They might need to start producing more parts on trailing fabs in order to get their capex down.

Keep in mind, integrated graphics was on a separate chip in the past, on older,fully depreciated fabs. now its on the leading edge fab which costs a lot more yet Intel has not been able to get any more money for the parts.

Just to give you an idea, continuously spending 12B capex means eventually depreciation will be 12B as well which means, depreciation alone is $40 per processor

What do those smartphone and tablet parts sell for again?

Thank you Pablo. It might be rough on the edge, but shows the need for action.

I know a lot of people dont like to hear talk like that.

What makes Intel a very profitable company, is that they over the years have been very agile in adapting for changes.

And they will do here to, but this is the worst strategic situation they have been in for years. And again reflected in share value.

The days when expensive x86 is needed is soon gone. A15 and Windows 8 is the start of the end.

The consolidation is accelerating, and Intel is fighting with Apple brand and Samsung production and huge vertical and horisontal span, be it products or technology. Add. everyone and his brother is buying from TSMC.

The ARM cost we get is often blown out of proportions. Zakate e350/450 production cost ex packaging was 9usd for aprox 70mm2 on 40nm TSMC - almost 2 years ago. I wouldnt be surpriced if you get a dualcore A15 on 28nm for the same H1 2013. Any new numbers on that?
Add. 2-3 usd for packaging and logistics. I dont remember ARM licensing, but is it even double digic %? - And all that is sans taking the risk.

Lets see what Intel does,- if they react and change plans now, it just show they are very much alive.
But dont count on your old desktop SB 2500k is rendered useless for highend gaming even in the next 6 years

 

[Fjodor2001]

http://forums.anandtech.com/showpost.php?p=34291877&postcount=55

What's the point of that? Then why not use a socket for the CPU to begin with as previously?

Why solder the CPU on a PCB, and then place the PCB in a socket on the motherboard?

 

[pablo87]

Thanks for the information.

http://seekingalpha.com/article/928...-2012-results-earnings-call-transcript?page=1 <---Transcript from Intel 2012 Q3 conference call.

I didn't see specific mention of 32nm equipment being replaced with 14nm. (The term "older generation technologies" was used instead, but Its not clear what node or nodes that refers to.)

Stacy Smith does mention too much 22nm inventory is a problem. (see 2nd quote). And if you look at the first part of the third quote I provided I just wonder if *some* 22nm equipment is getting replaced or converted to 14nm?

Regarding 32nm, I do know that node will still be used for Haswell and Broadwell PCH so keeping some of those fabs open will be part of Intel's plan.

Some quotes from Stacy Smith about 14nm:







Some interesting commentary (from the Intel 2012 Q3 conference call transcript) on re-using 22nm equipment for 14nm:

I heard 32, my mistake. Lots of discussion on capex, that's for sure.

 

[piasabird]

If you have a virtual monopoly you can do whatever you want. Of course the EU will probably just investigate them and lay a few fines on them. That is just part of doing business.