• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Panasonic / Intel Fab announcment

The question i have is what does Panasonic need 14nm capacity for? I cant imagine most of their consumer electronics need it. They do have a 4k action cam using internal soc's but the volume on that can't justify intel working with them. anybody have any thoughts?
 
Nice. Really exciting to see other companies being able to utilize Intel's enormous process advantage so Intel's lead won't be restricted to only Intel's chips anymore, also a bit like the Rockchip deal.
 
ShintaiDK said:
Seems like another customer that doesnt compete in any way with Intel. The trend stays the same.
Click to expand...

I guess that'd mean there will be no successor for Intel's Berryville - although that was probably obvious at this point. Not sure if anyone used it in the first place.
 
witeken said:
Nice. Really exciting to see other companies being able to utilize Intel's enormous process advantage so Intel's lead won't be restricted to only Intel's chips anymore, also a bit like the Rockchip deal.
Click to expand...

wasn't the rockchip deal that rockchip uses baytrail cores and integrate into a budget SOC manufactured by TSMC but labeled as an Intel product?
Complete opposite from the Panasonic deal in that case.
 
teejee said:
wasn't the rockchip deal that rockchip uses baytrail cores and integrate into a budget SOC manufactured by TSMC but labeled as an Intel product?
Click to expand...

Yes, but there's allegedly plans to eventually transition this to something fabbed by Intel. Why Intel wants Rockchip's partnership on this is something that still isn't totally clear to me.

teejee said:
Complete opposite from the Panasonic deal in that case.
Click to expand...

Well, we don't know if Panasonic is going to be using x86 cores or other Intel IP here. My initial guess would be that they aren't, but I only say that pretty weakly.

It looks like Japanese IDMs have been trying to figure out what to do with their fabrication for a long time now. That's one of the best markets for Intel to start picking off foundry customers.
 
teejee said:
wasn't the rockchip deal that rockchip uses baytrail cores and integrate into a budget SOC manufactured by TSMC but labeled as an Intel product?
Complete opposite from the Panasonic deal in that case.
Click to expand...

Yes, but the overall trend here is that Intel's becoming more open to cooperate with other companies. Who thought in 2012 that Intel would become a foundry or license their x86 IP to be sold by Rockchip?
 
My only question is will the be an "official" or "unofficial" separation when intel can use their fabs first vs other foundry customers.

Because intel is making the foundry, and they in theory has better communication with their cpu team than outside companies they will get a leg up and have a faster turn around on cpu design, synthesis, testing, quality control etc. They can figure out the process of the node even before the node is "commercial ready." This is the unofficial separation.

An officially separation would be, at they same time they can add in a delay to other customers that intel processors get the node for 6 months or 12 months before other foundry customers. That or Intel will not have to pay the "newest node" premium they may charge other customers.

My best is on unofficial separation, but I am really not sure. The answer is whatever makes intel the most money, or intel is convinced will make them the most money but I am not so sure.
 
I don't think Intel discloses such information. I guess Intel's chip business is still the same and they'll get highest priority. It's surely possible that other companies' chips will be released somewhat later. But even with a year delay, it would still a great deal.
 
witeken said:
Who thought in 2012 that Intel would become a foundry or license their x86 IP to be sold by Rockchip?
Click to expand...

Let's see - in 2009 Intel announced they'd license current Atom cores to TSMC (but that never went anywhere), and in 2010 Intel announced Achronix parts would be made on their 22nm process. By 2012 probably a lot of people were thinking this.
 
Roland00Address said:
My only question is will the be an "official" or "unofficial" separation when intel can use their fabs first vs other foundry customers.
Click to expand...

I'm guessing that so long as the other companies are only making chips that don't compete with anything Intel is doing, Intel wouldn't have a problem with them using their cutting edge fab processes.

However, if there's some market overlap, I imagine that either you won't get to use it or you'll have to wait until Intel has moved on to something else and they're just looking for customers to help recover costs.

While working out kinks in a new process, having chips with large die size just increases the likelihood of defects. If Intel can find someone who wants to make some really small chips where there won't be as much of a hit to the yield percentage and the customer doesn't directly compete with Intel that is willing to pay for the process while Intel is still refining it, why would Intel turn down that opportunity?
 
liahos1 said:
do tv's really need leading edge though? And how many smart tv's could panasonic really sell. maybe panasonic makes ia based tablet socs and goes merchant haha😕
Click to expand...

Its not just for TVs. 😉

But you be surprised. The ARM part in our slim Samsung TV is scorching hot on the backside. To add, its not exactly fast either. And then there is the TV tuner and decoder as well.
 
Last edited:
14nm fabs cost lots more than 22nm fabs. Intel claims significant reductions in cost per transistor @ 14nm. The only way that works is if each 14nm fab produces many more transistors. The transistors have to go somewhere and the desktop market is not smoking hot. Expect lots more of this in diverse markets. Under Paul's "crawl,walk run" foundry business plan, this looks like a fast crawl. Brian wants to run. Many more market opportunities may be targeted over time.

Is the quality of the design tool-set Intel provides to foundry customers comparable to the TSMC tool-set? TSMC should have better tools today. That may be the key to the "walk run" parts of Intel's plan.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top